# Release notes

Check out the latest updates and enhancements to Memgraph's products. Discover
what's new and understand the pivotal shifts that might affect your experience.
Dive into detailed listings of each version's changes, ensuring you stay
informed and prepared.

- [Check latest release notes](#-latest-release)

## 🚨 Important updates

New releases might affect your existing code, queries or configuration.
Ensure they're updated to align with the latest updates and changes.

> **Info**
>
> To ensure compatibility and stability, we urge you to **recompile your modules
> with each Memgraph update**. While an old build might still work, skipping
> recompilation could lead to unexpected crashes — better to be safe than
> troubleshoot in production.
>
> For step-by-step instructions on upgrading standalone instances or
> high-availability clusters, see the [Upgrades](https://memgraph.com/docs/database-management/upgrades)
> guide.

## 🚀 Latest release

### Memgraph v3.13.1 - September 14th, 2026

#### ✨ New features

- Added experimental, Enterprise-only FIPS 140-3 support: a new 
  `memgraph/memgraph:<version>-fips` Docker image bundling the validated OpenSSL
  FIPS provider, and a `--fips-mode=true` flag that restricts password hashing 
  to PBKDF2-SHA256 and TLS to 1.2 or higher.
  [#4664](https://github.com/memgraph/memgraph/pull/4664)

#### 🐞 Bug fixes

- Fix stale `MemoryRes` value in `SHOW METRICS INFO`, caused by this metric only 
  being updated when metrics were scraped from the HTTP metrics endpoint.
  [#4793](https://github.com/memgraph/memgraph/pull/4793)
- Fixed `SHOW METRICS INFO` failing with "Metrics registry has no database with 
  UUID" on a database that recovered from a snapshot written under a different 
  UUID, such as a replica recovering from its main. 
  [#4805](https://github.com/memgraph/memgraph/pull/4805)

### Lab v3.13.1 - September 11th, 2026

### Memgraph v3.13.0 - September 9th, 2026

#### ⚠️ Breaking changes

- The APOC compatibility names `apoc.convert.fromJsonList` and
  `apoc.convert.toJson` now resolve to the C++ `convert` module
  (`convert.from_json_list` / `convert.to_json`) instead of the Python
  `json_util` implementations. Output can differ — especially for nodes,
  relationships, paths, points, and temporals. Call `json_util.*` explicitly if
  you need the previous Python behavior, or update consumers for the new
  structured / ISO-8601 JSON forms.
  [#4443](https://github.com/memgraph/memgraph/pull/4443)
- The `COORDINATOR` privilege was removed — it never gated any operation. It is
  no longer accepted in `GRANT` / `DENY` / `REVOKE` statements, which now fail
  with a syntax error, and it is no longer reported by `SHOW PRIVILEGES` or
  included in `GRANT ALL PRIVILEGES`. Remove any `GRANT COORDINATOR` /
  `DENY COORDINATOR` / `REVOKE COORDINATOR` statements from provisioning scripts
  and update tooling that compares against the full privilege list. No data
  migration is required: a stale grant in an existing deployment is never
  reported and never checked. The new `COORDINATOR_READ` and `COORDINATOR_WRITE`
  privileges are not a replacement — they are coordinator-only and are rejected
  on data instances.
  [#4399](https://github.com/memgraph/memgraph/pull/4399)
- Connecting to a coordinator with a username and password is no longer always
  accepted. Once `--auth-module-mappings` is set, the enterprise license is
  valid, **and** a role holding `COORDINATOR_WRITE` exists, basic authentication
  is denied in favor of SSO. Until all three hold, the previous passthrough
  behavior is unchanged.
  [#4399](https://github.com/memgraph/memgraph/pull/4399)
- Inside a bare `CALL {}` subquery, a pattern that mentions an outer variable
  without importing it now opens a fresh binding instead of silently reusing
  the caller’s. Queries that relied on the old leak return different results
  with no error. Import explicitly with `CALL (n) {`, `CALL (*) {`, or a
  leading `WITH n`. Some invalid pattern-comprehension shapes that used to
  abort the server now raise a clear query error instead (create+reference in
  the same clause, reusing a bound relationship, variable-length expand from a
  node written in the same query part).
  [#4487](https://github.com/memgraph/memgraph/pull/4487)
- In `path.expand` / `path.expand_config` / `path.subgraph_*` relationship
  filters, `<TYPE>` now means incoming `TYPE` (the old reciprocal-pair reading
  is gone). `*` in a label filter is a wildcard, not a literal label. Unknown
  config keys are errors instead of being ignored. Calls that relied on the
  old readings can return different results or fail without a code change.
  [#4530](https://github.com/memgraph/memgraph/pull/4530)
- `SHOW INSTANCES` is now strongly consistent and always answered by the leader
  coordinator. A follower that cannot reach the leader no longer falls back to
  reporting the cluster from its own local Raft state with `unknown` health —
  it returns an empty result set with a `LeaderNotReachable` warning
  notification instead. Tooling that parses `SHOW INSTANCES` should treat an
  empty result as "cluster state unknown, retry" rather than "no instances
  registered". A data instance that is down now reports the role recorded in the
  Raft log (`main` / `replica`) instead of an unknown role.
  [#4492](https://github.com/memgraph/memgraph/pull/4492)
- `TERMINATE TRANSACTIONS` now requires transaction ids to parse in full. Ids
  with trailing characters previously terminated the transaction matching the
  numeric prefix, and unparseable ids were reported back as an attempt on
  `18446744073709551615`; both now raise an error.
  [#4534](https://github.com/memgraph/memgraph/pull/4534)
- `size()`, `substring()`, `left()`, and `right()` now count characters, not
  bytes. For non-ASCII text, lengths and cut positions can change (for example
  `size('中')` is `1`, not `3`). Recheck any stored results that depended on
  the old byte offsets.
  [#4586](https://github.com/memgraph/memgraph/pull/4586)
- A number that starts with `0` is now octal, so `09` and `018` are parse
  errors instead of `9.0` and `18.0`. Drop the leading zero for decimal, or
  keep it for octal (`010` is still `8`).
  [#4588](https://github.com/memgraph/memgraph/pull/4588)
- A `RETURN` inside `EXISTS { ... }` now affects the answer. Queries that
  relied on the body `RETURN` being ignored may return different results after
  upgrading.
  [#4597](https://github.com/memgraph/memgraph/pull/4597)
- Whole-number strings that do not fit in a 64-bit integer now raise an error
  in `toInteger()` instead of wrapping to a wrong value. Use `toIntegerOrNull()`
  if you want `null` for those inputs.
  [#4589](https://github.com/memgraph/memgraph/pull/4589)
- `single()` returns `null` when the predicate is `null` for any list element,
  instead of `true`. Queries that treated a null predicate as a match will see
  different results. Wrap with `coalesce(single(...), false)` if those rows
  should be treated as not matching.
  [#4590](https://github.com/memgraph/memgraph/pull/4590)
- `split('', ...)` returns `['']` rather than an empty list, so a non-null
  string always yields at least one field. Update callers that assumed splitting
  an empty string produced no fields.
  [#4591](https://github.com/memgraph/memgraph/pull/4591)
- The default value of `--metrics-format` changed from `JSON` to `OpenMetrics`.
  A deployment that relied on the JSON metrics endpoint without setting the flag
  now serves OpenMetrics after upgrading. Either update your monitoring
  dashboards, as described in [Migrating from JSON to
  OpenMetrics](https://memgraph.com/docs/database-management/monitoring#migrating-from-json-to-openmetrics),
  or set `--metrics-format=JSON` explicitly. The deprecated JSON endpoint
  remains available for several releases.
  [#4678](https://github.com/memgraph/memgraph/pull/4678)
- The `cross_database` module (and the deprecated `migrate.*` aliases) now
  requires Memgraph Enterprise. Community deployments that called these
  procedures need an Enterprise license after upgrading.
  [#4726](https://github.com/memgraph/memgraph/pull/4726)

#### 🐞 Bug fixes

- Fixed a query planning crash when `extract()` was used inside `WHERE` (for
  example `WHERE 'x' IN extract(v IN n.list | v.field)`). Those queries now
  run as written.
  [#4081](https://github.com/memgraph/memgraph/pull/4081)
- Identifiers may now include currency symbols such as `$` anywhere except as
  the first character (for example `price$`), matching openCypher. Whitespace
  between `$` and a parameter name (for example `$ 1`) is also handled
  correctly. Previously both cases could confuse the query stripper and crash
  the server.
  [#4283](https://github.com/memgraph/memgraph/pull/4283)
  [#4285](https://github.com/memgraph/memgraph/pull/4285)
- Text and vector search now respect fine-grained label and property
  permissions. Hits the caller cannot read are filtered out, and
  `text_search.aggregate` / `text_search.aggregate_edges` return an error unless
  every indexed property is readable.
  [#4316](https://github.com/memgraph/memgraph/pull/4316)
- Running many one-off queries that differ only in large inlined values (for
  example embeddings) no longer lets the query cache grow without bound and
  exhaust memory. The cache is capped by `--query-ast-cache-max-size` (default
  `1000`; `0` disables it), and each entry also uses less memory.
  [#4334](https://github.com/memgraph/memgraph/pull/4334)
  [#4445](https://github.com/memgraph/memgraph/pull/4445)
- Fixed parallel index scans (`USING PARALLEL EXECUTION`) so a filter that
  evaluates to `null` returns no rows instead of treating the filter as absent
  and returning every matching edge or node.
  [#4406](https://github.com/memgraph/memgraph/pull/4406)
- The `collections.*` functions and procedures now accept `NULL` list and scalar
  arguments instead of failing argument validation. Each function returns a
  defined result (for example `collections.sum(null)` → `null`,
  `collections.sort(null)` → `[]`, `collections.contains(null, x)` → `false`),
  matching the compatibility-layer behavior.
  [#4416](https://github.com/memgraph/memgraph/pull/4416)
- Fixed stale coordinator entries in the HA routing table and a possible
  coordinator crash on `REMOVE COORDINATOR`. Coordinators now deduplicate
  entries when applying Raft logs and snapshots.
  [#4444](https://github.com/memgraph/memgraph/pull/4444)
- Fixed `path.subgraph_all` and `path.subgraph_nodes` so their config options
  work as expected.
  [#4447](https://github.com/memgraph/memgraph/pull/4447)
- Under heavy concurrent load, operations that need exclusive access to the
  database could hang indefinitely while shared readers kept arriving. Those
  exclusive waiters now take priority so the hang no longer happens.
  [#4450](https://github.com/memgraph/memgraph/pull/4450)
- Fixed concurrent `SET STORAGE MODE` handling so mode switches are safe under
  load.
  [#4451](https://github.com/memgraph/memgraph/pull/4451)
- Fixed a crash when `LOAD PARQUET` hit invalid values (for example a time
  outside Memgraph’s supported range). The query now fails with an error
  instead of taking down the server.
  [#4454](https://github.com/memgraph/memgraph/pull/4454)
- Fixed `LOAD PARQUET` silently storing wrong timestamps or durations when
  converting large temporal values from Parquet. Those cases now fail with a
  clear query error instead of corrupting data.
  [#4455](https://github.com/memgraph/memgraph/pull/4455)
- Fixed a rare crash in garbage collection under heavy concurrent writes when
  indexes or unique constraints are defined.
  [#4475](https://github.com/memgraph/memgraph/pull/4475)
- Fixed pattern comprehensions in `WITH … WHERE` / `ORDER BY` so they correlate
  with the bound variable instead of scanning the whole graph; fixed `CALL {}`
  overwriting outer variables it did not import; and fixed related wrong
  results for named paths and `WITH` + `ORDER BY` + `WHERE`.
  [#4487](https://github.com/memgraph/memgraph/pull/4487)
- Fixed a crash when multiple sessions printed query plans at the same time
  (for example concurrent `EXPLAIN` / `PROFILE`, or query-plan logging).
  [#4489](https://github.com/memgraph/memgraph/pull/4489)
- Pattern comprehensions now see writes made earlier in the same query part.
  Reading a list like `[(p)-[:ACTED_IN]->(m) | m]` after a `SET` / `CREATE` no
  longer returns empty or wrong results. Expanding from a node that a preceding
  clause deleted now raises an error instead of counting relationships that no
  longer exist.
  [#4499](https://github.com/memgraph/memgraph/pull/4499)
- Fixed TTL sometimes never running after it was configured or enabled, so
  expired data is cleaned up as scheduled.
  [#4513](https://github.com/memgraph/memgraph/pull/4513)
- `SET COORDINATOR SETTING 'instance_health_check_frequency_sec' TO '0'` is now
  rejected with an invalid-argument error instead of crashing the coordinator
  later (for example on `REGISTER INSTANCE` or restart).
  [#4522](https://github.com/memgraph/memgraph/pull/4522)
- Fixed silent data loss when a WAL file other than the last one in the chain
  was damaged. Recovery used to apply the readable prefix of that file and then
  apply the following files in full, coming up with an incomplete dataset
  without reporting anything; on a replica the divergence never healed because
  the replica reported the later timestamp and main considered it caught up.
  Damage in a completed (finalized) WAL file is now fatal: the database fails
  recovery instead of starting stale, and a replica stops applying the rest of
  the chain. A WAL file that was still being written when the process crashed
  is unaffected and is still truncated to its last whole transaction. Recover
  such an instance with
  [`--storage-allow-recovery-failure`](https://memgraph.com/docs/database-management/configuration) plus
  `RECOVER SNAPSHOT`, or from a backup.
  [#4528](https://github.com/memgraph/memgraph/pull/4528)
- Fixed several `path` module issues: traversals honor query timeout /
  `TERMINATE TRANSACTIONS` / shutdown; deep expansions error instead of
  crashing; `path.create` no longer aborts on a missing `rel` key; and filter /
  hop-bound edge cases that produced wrong or empty results.
  [#4530](https://github.com/memgraph/memgraph/pull/4530)
- `TERMINATE TRANSACTIONS` no longer reports `killed: true` for a transaction
  the caller is not authorized to terminate. Such a match now reports
  `killed: false`, indistinguishable from an id that does not exist.
  [#4534](https://github.com/memgraph/memgraph/pull/4534)
- Fixed a crash when writing a property covered by a nested index if the value
  has nothing at the indexed path (for example `CREATE (:T {m: 1})` with an
  index on `:T(m.k)`).
  [#4550](https://github.com/memgraph/memgraph/pull/4550)
- `path.slice` clamps out-of-range `offset` / `length` instead of throwing, and
  `path.slice`, `path.elements`, and `path.combine` accept a null path. A
  negative `length` other than `-1` (to the end) now returns a subpath with no
  relationships rather than the rest of the path.
  [#4552](https://github.com/memgraph/memgraph/pull/4552)
- `*KSHORTEST|k` now returns up to `k` paths for each input row, not for the
  whole query. `|k` can also be an expression from the current row (for example
  `-[*KSHORTEST|n.k]->`). With an inlined edge property map, only matching
  paths count toward `k`.
  [#4559](https://github.com/memgraph/memgraph/pull/4559)
- Negative or empty `*KSHORTEST` depth bounds and a negative `|k` now raise a
  query error instead of returning no rows (or scanning the whole graph first).
  [#4559](https://github.com/memgraph/memgraph/pull/4559)
- `EXISTS { ... }` subquery bodies now behave as expected: bodies that always
  match evaluate to `true` instead of crashing the server, and `EXISTS` in
  `WITH` projections and `ORDER BY` no longer fails with a spurious aliasing
  error.
  [#4560](https://github.com/memgraph/memgraph/pull/4560)
- Replicas no longer time out during recovery steps that run long inside a
  single delta, such as populating an index, validating a constraint, loading a
  snapshot, clearing storage before a reset, or aborting an interrupted
  two-phase commit. A replica now reports progress on a fixed time interval for
  as long as the work keeps advancing, instead of after a fixed number of
  processed deltas, so the main no longer drops the connection mid-build and
  leaves the replica unable to converge. A replica that stops advancing stops
  reporting, so a genuinely stuck replica is still caught by the main's timeout.
  As a consequence, the `deltas_batch_progress_size` coordinator setting **no
  longer has any effect**; it stays readable and settable so existing
  configurations keep working across an upgrade, and will be removed in a future
  release. [#4557](https://github.com/memgraph/memgraph/pull/4557)
- Filters on a variable imported into a scoped `CALL (v) { ... }` subquery now
  use a label-property or `id()` index instead of scanning every node or edge
  once per input row. `MATCH ()-[e:TYPE]->() WHERE id(e) = ...` no longer
  returns an edge of the wrong type.
  [#4561](https://github.com/memgraph/memgraph/pull/4561)
- Variables imported by `CALL (v1, v2) { ... }` and `CALL (*) { ... }` now stay
  bound after a `WITH` inside the subquery. Previously an inner `WITH` that
  omitted them could return wrong rows, overwrite the outer variable, or
  `CREATE`/`MERGE` a new node instead of using the imported one.
  [#4578](https://github.com/memgraph/memgraph/pull/4578)
- Fixed crashes and unbounded index growth in `IN_MEMORY_ANALYTICAL` when
  indexes are enabled. Index entries are now removed before the nodes or edges
  they point to are deleted.
  [#4580](https://github.com/memgraph/memgraph/pull/4580)
- `reverse()` now reverses characters, not bytes, so accented text, Chinese,
  and emoji come back readable instead of as broken UTF-8.
  [#4585](https://github.com/memgraph/memgraph/pull/4585)
- `size()`, `substring()`, `left()`, and `right()` now work on whole
  characters, so cuts no longer split multi-byte text into invalid UTF-8.
  [#4586](https://github.com/memgraph/memgraph/pull/4586)
- Integer literals with a leading zero are no longer treated as floating-point
  values, so they get a proper integer range check instead of a silent
  approximate double.
  [#4588](https://github.com/memgraph/memgraph/pull/4588)
- `toInteger()` now converts whole-number strings exactly. Values near the
  64-bit limits no longer round to a neighbouring integer (for example the
  largest integer no longer comes back as the smallest).
  [#4589](https://github.com/memgraph/memgraph/pull/4589)
- An invalid timezone now raises a clear argument error instead of "An
  unknown exception occurred". UTC offsets such as `+01:00` and `Z` are
  accepted as well as timezone names.
  [#4592](https://github.com/memgraph/memgraph/pull/4592)
- A `RETURN` inside `EXISTS { ... }` is now planned instead of ignored, so
  `DISTINCT`, `SKIP`, `LIMIT`, and aggregation in the body affect the answer.
  Results can change without an error — recheck any query that relied on the
  body `RETURN` being ignored.
  [#4597](https://github.com/memgraph/memgraph/pull/4597)
- Replication and coordination now log cluster state changes at `INFO` instead
  of `TRACE`, so you can see failover, role, and leadership changes without
  enabling per-delta trace logs. The default log level stays `WARNING`.
  [#4599](https://github.com/memgraph/memgraph/pull/4599)
- Query log masking no longer leaks part of a secret when the value contains
  quotes or is missing a closing quote. Those values are now fully redacted in
  logs.
  [#4629](https://github.com/memgraph/memgraph/pull/4629)
- `EXPLAIN` no longer shows a `Limit` above an `EXISTS { ... }` or
  `COUNT { ... }` subquery branch. That operator did not actually limit the
  branch, so the plan looked wrong. Query results are unchanged.
  [#4630](https://github.com/memgraph/memgraph/pull/4630)
- Type constraints now accept zoned datetimes whose timezone is a numeric UTC
  offset, such as `+01:00`, instead of rejecting them as the wrong type.
  [#4635](https://github.com/memgraph/memgraph/pull/4635)
- Near the memory hard limit, Memgraph no longer crashes by refusing allocations
  the runtime needs for itself. Query allocations are still limited as before.
  [#4646](https://github.com/memgraph/memgraph/pull/4646)
- Repeated runs of variable-length expands that return distinct vertices (for
  example with `DISTINCT`) are faster, because those queries no longer replan
  on every execution.
  [#4650](https://github.com/memgraph/memgraph/pull/4650)
- Setting a list property that is also covered by a vector index no longer
  crashes the server when the node already has a typed-list constraint. The
  value is checked like any other list.
  [#4653](https://github.com/memgraph/memgraph/pull/4653)
- List comprehensions whose `WHERE` predicate is `NULL` for an element now
  omit that element and keep the rest of the list, instead of returning
  `NULL` for the whole result. A `WHERE` predicate that is neither a boolean
  nor `NULL` is rejected with an error.
  [#4665](https://github.com/memgraph/memgraph/pull/4665)
- `LOAD PARQUET` and `LOAD JSONL` accept URLs of any length, including long
  signed URLs. A download in progress can be stopped with
  `TERMINATE TRANSACTIONS`, and failures that cannot succeed on retry are
  reported as client errors rather than transient ones.
  [#4672](https://github.com/memgraph/memgraph/pull/4672)
- Lists that hold large integers are compared and sorted correctly. Values that
  differed only in the high bits no longer compared equal.
  [#4712](https://github.com/memgraph/memgraph/pull/4712)

#### 🛠️ Improvements

- Memgraph is now built with a distro-agnostic toolchain, so DEB and RPM
  packages install on a wider range of Linux distributions that meet
  GLIBC >= 2.31 and Python >= 3.10.
  [#3587](https://github.com/memgraph/memgraph/pull/3587)
- `node2vec` and `node2vec_online` now use a C++ word2vec implementation instead
  of the unmaintained Python `gensim` package. Usage is unchanged, expect
  faster runs.
  [#4319](https://github.com/memgraph/memgraph/pull/4319)
- MAGE is now built and packaged from the same unified CMake / Conan tree as
  Memgraph. Build it with `build.sh --mage on` or `--mage only`; the old
  `mage/setup` Python script is removed. Only affects people building MAGE from
  source — prebuilt packages are unchanged.
  [#4348](https://github.com/memgraph/memgraph/pull/4348)
- `build.sh --target` now accepts several targets in one invocation (for
  example `--target memgraph memgraph__unit`), so you can build multiple
  targets with a single configure step instead of one run per target.
  [#4404](https://github.com/memgraph/memgraph/pull/4404)
- Queries that use query-module functions or `CALL` procedures are now
  plan-cached instead of being re-parsed and re-planned on every run. Module
  reloads stay safe; only module-dependent queries are invalidated, and the
  next execution picks up updated code.
  [#4461](https://github.com/memgraph/memgraph/pull/4461)
  [#4482](https://github.com/memgraph/memgraph/pull/4482)
- In in-memory transactional mode, `CREATE INDEX` / `CREATE CONSTRAINT` no
  longer stall behind garbage collection (and GC no longer waits on those DDL
  statements). Under index-heavy workloads this removes GC-correlated latency
  spikes on index and constraint creation.
  [#4468](https://github.com/memgraph/memgraph/pull/4468)
- `SHOW TRIGGERS` and `DROP TRIGGER` no longer take a storage write lock.
  `CREATE TRIGGER` uses a read accessor only. Clients that poll
  `SHOW TRIGGERS` often (for example Memgraph Lab) no longer contend with
  writers for that reason.
  [#4495](https://github.com/memgraph/memgraph/pull/4495)
- Connecting Lab is less expensive under load: its connect-time health-check
  queries no longer open a storage transaction or pin a tenant.
  [#4496](https://github.com/memgraph/memgraph/pull/4496)
- The self-contained Memgraph+MAGE offline installer is quieter (installer noise
  goes to a log file) and is published with the other release artifacts on the
  [download hub](https://memgraph.com/download).
  [#4500](https://github.com/memgraph/memgraph/pull/4500)
- Garbage collection is faster on databases with many indexes: it only visits
  indexes that recent writes could have touched.
  [#4508](https://github.com/memgraph/memgraph/pull/4508)
- Recovery from WAL files is roughly 35% faster. Each WAL file now records its
  timestamp range and transaction count in its header, so deciding which files
  to recover from no longer requires parsing every file end to end, and
  replaying a file parses it once instead of twice. This also speeds up replica
  recovery and the WAL cleanup that runs after every snapshot.
  [#4528](https://github.com/memgraph/memgraph/pull/4528)
- Path config filters are stricter: empty or bare filter entries, `*` in a
  relationship filter, and blank sequence steps are rejected instead of
  producing wrong results. On `path.subgraph_nodes` / `path.subgraph_all`,
  `bfs` and `uniqueness` are accepted and ignored (those procedures are always
  breadth-first, one visit per node).
  [#4530](https://github.com/memgraph/memgraph/pull/4530)
- Queries that filter with `IN` on an indexed label and property (for example
  `MATCH (x:L) WHERE x.p IN ["a", "b"]`) now get accurate cardinality estimates,
  so the planner picks better index plans for those filters.
  [#4538](https://github.com/memgraph/memgraph/pull/4538)
- Writing and loading a snapshot no longer fills the operating system's file
  cache, so more memory stays free for queries while it happens. This works
  automatically; `--storage-snapshot-writeback-window-mib` tunes how much of a
  snapshot is cached before being written to disk (`0` restores the old
  behaviour).
  [#4546](https://github.com/memgraph/memgraph/pull/4546)
- Variable-length expands that return distinct vertices (for example with
  `DISTINCT`) no longer walk every edge-unique path. Memgraph can use a
  pruning breadth-first search instead, which is much faster on wide or dense
  graphs.
  [#4558](https://github.com/memgraph/memgraph/pull/4558)
- `CONTAINS`, `STARTS WITH`, and `ENDS WITH` on an indexed label and property
  now use that index instead of scanning every node.
  [#4562](https://github.com/memgraph/memgraph/pull/4562)
- Replicating edge deletions is much faster on replicas. Large deletes that
  were previously quadratic on the replica now scale with the number of edges
  deleted, as they do on MAIN.
  [#4564](https://github.com/memgraph/memgraph/pull/4564)
- `WHERE NOT NOT n.prop = 42` (and similar double negation) now uses a property
  index instead of scanning the label. Results are unchanged.
  [#4594](https://github.com/memgraph/memgraph/pull/4594)
- In `IN_MEMORY_ANALYTICAL`, `CREATE INDEX` and `DROP INDEX` no longer block
  read queries. You can also build several supported indices in parallel, while
  writes are still held off until the build finishes.
  [#4633](https://github.com/memgraph/memgraph/pull/4633)
- Fixed unnecessary work on property writes when fine-grained property
  permissions are not configured.
  [#4639](https://github.com/memgraph/memgraph/pull/4639)
- Indexed `STARTS WITH` filters no longer re-check the prefix after the index
  scan, so those queries do less work per row. Results are unchanged.
  [#4643](https://github.com/memgraph/memgraph/pull/4643)
- Commits on MAIN with replicas are faster: encoding work for replicas runs in
  parallel with writing the local WAL, so commit time no longer grows as
  steeply with the number of registered replicas.
  [#4645](https://github.com/memgraph/memgraph/pull/4645)
- Indexed `CONTAINS`, `ENDS WITH`, and `REGEX` filters are faster when many
  nodes share the same property value: after one value fails the filter, the
  scan skips the rest of that value instead of checking every duplicate.
  [#4647](https://github.com/memgraph/memgraph/pull/4647)
- Aggregations such as `count()`, `sum()`, and `avg()` are faster, especially
  without a grouping key, because they do less work per input row. Results are
  unchanged.
  [#4655](https://github.com/memgraph/memgraph/pull/4655)
- `path.expand_config`, `path.expand`, `path.subgraph_nodes`, and
  `path.subgraph_all` are substantially faster, especially with selective
  filters or a single relationship direction.
  [#4727](https://github.com/memgraph/memgraph/pull/4727)
- Cluster management queries forwarded from a follower coordinator to the leader
  now have [explicit RPC
  timeouts](https://memgraph.com/docs/clustering/high-availability/how-high-availability-works#rpc-timeouts),
  so a Bolt session can no longer block indefinitely against a leader that is
  reachable but stuck. The same applies to MAIN-to-REPLICA system transaction
  messages (auth, multi-tenancy, parameters), which now time out after 10s and
  defer to system recovery. The unused `EnableWritingOnMainRpc` and its
  Prometheus counters were removed.
  [#4399](https://github.com/memgraph/memgraph/pull/4399)
- `YIELD LEADERSHIP` and `SHOW COORDINATOR SETTINGS` can now be run on any
  coordinator — followers forward them to the leader instead of failing or
  answering from local state. When no leader can serve a cluster query,
  Memgraph now says so explicitly: read queries (`SHOW INSTANCES`, `SHOW
  COORDINATOR SETTINGS`, `SHOW REPLICATION LAG`) return no rows with a
  `LeaderNotReachable` warning, `SHOW REPLICATION LAG` additionally reports why
  the lag is unavailable via `ReplicationLagUnavailable` (no current main, main
  unresponsive, leader still taking over, stale leader view), and
  `ADD COORDINATOR`, `REMOVE COORDINATOR` and `UPDATE CONFIG` fail with a clear
  "coordinator is not a leader" error instead of a misleading one.
  [#4492](https://github.com/memgraph/memgraph/pull/4492)

#### ✨ New features

- Added global vertex-property indexes with `CREATE GLOBAL INDEX ON :(prop)`
  (and `DROP GLOBAL INDEX ON :(prop)`). They speed up property lookups across
  all nodes when there is no shared label, for example
  `MATCH (n) WHERE n.uid = '123' RETURN n`.
  [#4353](https://github.com/memgraph/memgraph/pull/4353)
- Added `collections.disjunction`, `collections.subtract`, and
  `collections.duplicates` to the MAGE `collections` module for set-style list
  difference, subtraction, and finding repeated values.
  [#4415](https://github.com/memgraph/memgraph/pull/4415)
- Added `map.get` and `map.merge_list` to the MAGE `map` module. Existing
  helpers such as `map.merge` and `map.set_key` now treat a `null` map as empty
  (and a `null` key as a no-op) instead of failing, and accept nodes or
  relationships by using their property maps.
  [#4417](https://github.com/memgraph/memgraph/pull/4417)
- Added `text.compare_cleaned(text1, text2)` to the MAGE `text` module. It
  compares two strings for equality after keeping only ASCII letters and digits
  (lowercased) and dropping everything else; `NULL` arguments return `false`.
  [#4435](https://github.com/memgraph/memgraph/pull/4435)
- C++ query-module authors can return a value whose type is only known at
  runtime with a single `Result::SetValue(value)` call, instead of switching
  on the value's type to pick a typed overload.
  [#4436](https://github.com/memgraph/memgraph/pull/4436)
- Added `convert.from_json_map`, `convert.from_json_list`, `convert.to_map`, and
  `convert.to_json` for parsing and serializing JSON, with an optional `path`
  selector on the from_* helpers and structured output for graph and temporal
  types from `to_json`.
  [#4443](https://github.com/memgraph/memgraph/pull/4443)
- Added `search.node` and `search.node_all` procedures that find nodes by a
  `{label: property}` map, a comparison operator, and a value. A property list
  per label is OR-ed; `search.node` de-duplicates by node while
  `search.node_all` keeps duplicates. An existing label-property index is used
  automatically when available.
  [#4460](https://github.com/memgraph/memgraph/pull/4460)
- `--coordinator-id=0` is now a valid coordinator id (previously rejected).
  [#4483](https://github.com/memgraph/memgraph/pull/4483)
- `CREATE RANGE INDEX FOR ... ON ...` now works for nodes and relationships and
  creates Memgraph’s usual property index.
  [#4486](https://github.com/memgraph/memgraph/pull/4486)
- [SSO authentication on high availability
  coordinators](https://memgraph.com/docs/clustering/high-availability/coordinator-authentication)
  (Enterprise). Coordinators authenticate OIDC, SAML and Kerberos connections
  against a Raft-replicated set of roles and enforce two privileges,
  `COORDINATOR_READ` and `COORDINATOR_WRITE`. Manage them with `CREATE ROLE`,
  `DROP ROLE`, `SHOW ROLES`, `GRANT` / `REVOKE COORDINATOR_READ |
  COORDINATOR_WRITE | ALL PRIVILEGES`, and `SHOW PRIVILEGES FOR ROLE` on any
  coordinator — follower queries are forwarded to the leader. Basic
  authentication stays open until a `COORDINATOR_WRITE` role exists, so no
  coordinator restart is needed to roll SSO out, and it reopens if the license
  becomes invalid so a coordinator can never lock every session out.
  [#4399](https://github.com/memgraph/memgraph/pull/4399)
- `SHOW VERSION` can now be run on coordinators, alongside the other read-only
  introspection queries such as `SHOW CONFIG` and `SHOW BUILD INFO`. It
  previously failed with `Coordinator can run only coordinator queries!`. On
  coordinators it is a `COORDINATOR_READ` query, so an SSO role with only
  `COORDINATOR_READ` granted can run it.
  [#4535](https://github.com/memgraph/memgraph/pull/4535)
- Added the `SHOW ROUTING TABLE` query, which shows the routing table a
  coordinator hands out to `bolt+routing` clients as `WRITE`, `READ` and `ROUTE`
  rows. It can only be run on a coordinator, is always answered from the
  leader's state, and returns an empty table if the leader cannot be contacted.
  [#4502](https://github.com/memgraph/memgraph/pull/4502)
- `EXISTS { ... }` and `exists((a)-[]->(b))` now work in `WITH` / `RETURN`
  projections, `WITH ... WHERE`, `ORDER BY`, and aggregate arguments, not only
  in a `MATCH` `WHERE`. You can write
  `WITH p, EXISTS { MATCH (p)-[:ACTED_IN]->() } AS hasActed` instead of
  rewriting it as an `OPTIONAL MATCH` plus a count.
  [#4504](https://github.com/memgraph/memgraph/pull/4504)
- Added `TERMINATE TRANSACTIONS "*"`, which terminates every transaction the
  caller is authorized to terminate across all databases, instead of naming ids
  one by one. The caller's own transaction is skipped, transactions it may not
  terminate are omitted from the result, and rows come back ordered by
  ascending transaction id. The wildcard cannot be combined with ids.
  [#4534](https://github.com/memgraph/memgraph/pull/4534)
- A data instance in a high availability cluster can now use the
  `IN_MEMORY_ANALYTICAL` storage mode for a bulk import: unregister the
  replicas, switch to analytical, import, switch back to
  `IN_MEMORY_TRANSACTIONAL` and register the replicas again. The switch back
  writes a snapshot of the imported data, archives the superseded durability
  files and restarts WAL numbering, and replica recovery detects that the
  imported data exists in no WAL file, so a re-registered replica catches up
  from that snapshot without wiping its data directory. Starting a data instance
  with `--storage-mode=IN_MEMORY_ANALYTICAL` remains forbidden. See [bulk import
  in analytical mode](https://memgraph.com/docs/clustering/high-availability/analytical-import).
  [#4510](https://github.com/memgraph/memgraph/pull/4510)
- Added property-value descriptions: attach a human-readable label to a
  specific property value with `SET DESCRIPTION ON PROPERTY … VALUE …`, resolve
  it in queries with `description(property, value)`, and manage them with the
  existing `SHOW` / `DELETE DESCRIPTION` commands.
  [#4526](https://github.com/memgraph/memgraph/pull/4526)
- Added `path.expand_config(start, config)` and extended the path config map
  with node-identity filters, `limit`, `uniqueness` (including `NODE_GLOBAL`),
  and repeating label/relationship sequences. `path.subgraph_nodes` and
  `path.subgraph_all` accept the new filter keys too.
  [#4530](https://github.com/memgraph/memgraph/pull/4530)
- Added `--storage-omit-vector-index-properties-on-return` (also a runtime
  database setting, default `false`). When enabled, properties backed by a
  vector index are omitted from whole nodes and relationships returned to the
  client, which cuts payload size for embeddings. Explicit access such as
  `n.embedding` still returns the value; dumps, WAL, and replication are
  unchanged.
  [#4556](https://github.com/memgraph/memgraph/pull/4556)
- `*KSHORTEST` now accepts a filter lambda, `-[*KSHORTEST (r, n | pred)]->`, so
  you can narrow the search while expanding instead of filtering whole paths
  afterwards.
  [#4559](https://github.com/memgraph/memgraph/pull/4559)
- `EXISTS { ... }` can now be used inside a `CASE` expression, including in the
  condition (for example `CASE WHEN EXISTS { ... } THEN ... END`).
  [#4596](https://github.com/memgraph/memgraph/pull/4596)
- Added `COUNT { ... }`, which counts the rows a subquery body produces (for
  example `RETURN COUNT { MATCH (p)-[:KNOWS]->(f) }`). Use it instead of
  building a list with a pattern comprehension and calling `size()`.
  [#4598](https://github.com/memgraph/memgraph/pull/4598)
- If a write cannot reach every SYNC replica but still commits on MAIN (and
  reachable replicas), the query no longer fails with an exception. A WARNING
  notification is returned instead, so the client result matches what was
  actually committed. Writes that are rolled back everywhere (for example
  STRICT_SYNC) still raise an error as before.
  [#4601](https://github.com/memgraph/memgraph/pull/4601)
- Added `COLLECT { ... }`, which returns a list built from a subquery body's
  single return column, evaluated once per outer row (for example
  `RETURN COLLECT { MATCH (p)-[:KNOWS]->(f) RETURN f.name }`). An empty match
  yields `[]` rather than `null`.
  [#4632](https://github.com/memgraph/memgraph/pull/4632)
- Added
  [`--storage-rocksdb-keep-log-file-num`](https://memgraph.com/docs/database-management/configuration#storage)
  (default `1000`), which caps how many RocksDB info log files are kept per
  RocksDB instance. Since every restart rolls the current info log into a new
  file, lowering it helps on deployments with constrained disk space.
  `--storage-rocksdb-info-log-level` and the new flag now apply to every RocksDB
  instance Memgraph opens — the key-value stores backing authentication,
  triggers, streams, settings and replication or high availability state — and
  not only to the `ON_DISK_TRANSACTIONAL` storage.
  [#4579](https://github.com/memgraph/memgraph/pull/4579)

### Lab v3.13.0 - September 9th, 2026

## Previous releases

### Memgraph v3.12.0 - July 15th, 2026

#### ⚠️ Breaking changes

- The deprecated high availability startup flags `--instance-down-timeout-sec`
  and `--instance-health-check-frequency-sec` have been removed after two release
  cycles. They were deprecated in 3.10 and ignored since then, but startup now
  fails if either flag is still set. Remove them from coordinator configs and
  use `SET COORDINATOR SETTING 'instance_down_timeout_sec' TO '<value>'` and
  `SET COORDINATOR SETTING 'instance_health_check_frequency_sec' TO '<value>'`
  instead. Values previously passed via the flags are not migrated automatically
  — re-apply any non-default settings after upgrade.
  [#4310](https://github.com/memgraph/memgraph/pull/4310)

#### ✨ New features

- Added property-based access control for nodes and relationships. `GRANT`,
  `DENY`, and `REVOKE` now accept a property list (`{prop, ...}` or `{*}` for all
  properties) alongside existing label and permission rules, so you can
  restrict which properties a user or role can read or write on matching nodes.
  [#4156](https://github.com/memgraph/memgraph/pull/4156)
- Added opt-in lightweight edges via `--storage-light-edge=true` to cut roughly
  24B of memory overhead per relationship when you need properties on edges.
  Edges are pool-allocated instead of stored in a dedicated skip list container,
  while traversals and pattern matching behave as before. Enable
  `--storage-enable-edges-metadata=true` if your workload resolves edges by
  internal ID.
  [#4228](https://github.com/memgraph/memgraph/pull/4228)
- Added `SUSPEND DATABASE` and `RESUME DATABASE` for multi-tenant Enterprise
  deployments. Suspend tears down a tenant's in-memory storage to a durable COLD
  shell to free RAM while keeping its data on disk; resume rebuilds it HOT with
  all data intact. Operations replicate across the cluster like `CREATE`/`DROP
  DATABASE`.
  [#4280](https://github.com/memgraph/memgraph/pull/4280)
- MAGE is now available as prebuilt RPM packages on CentOS 9 and CentOS 10, so
  you can install the graph algorithm library with your package manager instead
  of building from source. Install the matching `memgraph` package first, then
  the `memgraph-mage` RPM.
  [#4289](https://github.com/memgraph/memgraph/pull/4289)
- Added `text_search.fuzzy_phrase_search` and
  `text_search.fuzzy_phrase_search_edges` for typo-tolerant phrase matching on a
  single property. Queries use `data.<property>:<terms>` with ordered, adjacent
  terms and prefix matching on the last term, so `data.title:big bad wo` can
  match `big bad wolf` or `big bad world` when `fuzzy_distance` allows a few
  edits across the phrase.
  [#4304](https://github.com/memgraph/memgraph/pull/4304)
- Added a cluster-wide coordinator read-only setting synced across coordinators,
  so operators can enable global read-only mode from any coordinator in the
  cluster.
  [#4330](https://github.com/memgraph/memgraph/pull/4330)
- Added the `--storage-allow-recovery-failure` flag (default `false`) so an
  in-memory database that fails durability recovery on startup comes up in a
  `broken` state instead of crashing the whole process. A broken database is
  empty, leaves its on-disk files untouched, and rejects data queries; healthy
  databases keep working. Health is reported via a new `Health` column in `SHOW
  DATABASES` and a `health` row in `SHOW STORAGE INFO`. Recover a broken database
  in place with `RECOVER SNAPSHOT`, restore the data directory from a backup, or
  drop it; on a cluster, a broken replica database self-heals from the main while
  a broken main stays main (no automatic failover).
  [#4284](https://github.com/memgraph/memgraph/pull/4284)

#### 🛠️ Improvements

- WAL files are now protected with CRC32 checksums. The WAL header and each
  transaction carry their own checksum, which is verified during recovery and on
  replicas before the data is applied, so silent on-disk corruption is detected
  instead of being recovered. WAL files written by older versions remain
  recoverable without verification.
  [#4225](https://github.com/memgraph/memgraph/pull/4225)
- Text search and regex queries return results faster, especially regex searches
  with large match sets. Regex top-N queries now stop once the configured limit
  is reached instead of scanning every match, so broad patterns on large indexes
  complete much sooner.
  [#4286](https://github.com/memgraph/memgraph/pull/4286)
- `MATCH (n) WHERE id(n) IN <list>` and label property `IN <list>` filters now
  use indexed scans when the list is a parameter or expression, instead of
  scanning the full graph. Bulk id lookups and parameterized membership filters
  run much faster on large databases.
  [#4299](https://github.com/memgraph/memgraph/pull/4299)
- Label-property indexes for one or two properties now store entries inline
  instead of allocating a separate heap buffer per entry, reducing memory
  overhead for the most common index configurations.
  [#4308](https://github.com/memgraph/memgraph/pull/4308)
- Reduced query overhead when property-based access control is enabled by
  caching the fine-grained auth checker per session and skipping property
  permission checks when no property-level restrictions apply.
  [#4309](https://github.com/memgraph/memgraph/pull/4309)
- Production Docker images now ship with supply chain attestations and SPDX
  SBOMs, signed with cosign. You can verify image signatures and retrieve the
  SBOM for installed images to support security auditing and compliance
  workflows.
  [#4337](https://github.com/memgraph/memgraph/pull/4337)

#### 🐞 Bug fixes

- Fixed `replace()` causing an infinite loop and OOM when the search string is
  empty. Empty search now inserts the replacement at every character boundary
  with bounded allocation, so queries such as `replace("abc", "", "-")` return
  `"-a-b-c-"` instead of crashing the server.
  [#4269](https://github.com/memgraph/memgraph/pull/4269)
- Fixed Bolt 4.4 clients receiving routing tables from data instances and
  incorrectly defaulting the target database. Data instances now reject routing
  requests, and database selection over Bolt 4.4 works as expected.
  [#4270](https://github.com/memgraph/memgraph/pull/4270)
- Fixed user and role data being corrupted during rolling HA upgrades between
  Memgraph versions. Auth entities now carry a version field and are migrated
  on deserialize, so fine-grained access grants replicated from an older MAIN
  are preserved after failover to an upgraded REPLICA.
  [#4301](https://github.com/memgraph/memgraph/pull/4301)
- Fixed duplicated coordinator entries in the routing table when re-adding a
  coordinator that was the current leader.
  [#4311](https://github.com/memgraph/memgraph/pull/4311)
- Fixed negative replication lag values in the routing table that could produce
  invalid routing entries when re-adding a coordinator or during failover.
  [#4312](https://github.com/memgraph/memgraph/pull/4312)
- Fixed MAIN instances hanging on shutdown while snapshot or WAL replication is
  still in progress. Shutdown now aborts the replication client so the instance
  stops immediately instead of waiting for replication to finish.
  [#4314](https://github.com/memgraph/memgraph/pull/4314)
- Fixed replication failures during WAL recovery when file-bearing RPC messages
  were split across network reads. Consumed message data is now dropped from the
  buffer so files are not parsed twice.
  [#4317](https://github.com/memgraph/memgraph/pull/4317)
- Fixed a rare server crash under heavy concurrent write load caused by a
  heap-use-after-free between transaction commit and storage garbage collection.
  Committed deltas are now registered before the commit is marked finished, so
  fast-discard and GC no longer free memory still referenced by in-flight
  transactions.
  [#4323](https://github.com/memgraph/memgraph/pull/4323)
- Fixed edge property index scans scaling linearly with index size for bounded
  lookups. Scans now seek directly to the lower bound instead of walking from
  the head of the skiplist.
  [#4326](https://github.com/memgraph/memgraph/pull/4326)
- Fixed several planner bugs that caused wrong or missing rows for edge
  `id(e) IN [...]`, edge property `IN` lists with duplicates, `ORDER BY` over
  unwind-fed index scans, composite-index nullable ordering, and nested edge
  property filters.
  [#4336](https://github.com/memgraph/memgraph/pull/4336)
- Fixed snapshot and WAL transfer files not being cleaned up when the final RPC
  response fails to send. Files are now moved to the `.old` directory regardless
  of send success, preventing orphaned durability files on replicas.
  [#4347](https://github.com/memgraph/memgraph/pull/4347)
- Track memory reclamation when jemalloc's background dirty-page decay returns
  pages to the OS. This caused reported memory usage to exceed actual memory
  usage. [#4396](https://github.com/memgraph/memgraph/pull/4396)

### Lab v3.12.1 - August 28th, 2026

### Lab v3.12.0 - July 15th, 2026

### Memgraph v3.11.0 - June 17th, 2026

#### ⚠️ Breaking changes

- `--query-modules-directory` is now ignored on coordinator instances and
  query modules are no longer loaded there. The embedded Python runtime is also
  skipped on coordinators, so any Python-based modules that previously loaded
  (but were unusable) will no longer be initialized. The flag itself is still
  accepted so packaged defaults do not need to be overridden when starting a
  coordinator. Additionally, `storage.snapshot.interval` is no longer registered
  as a runtime setting on coordinators — `SHOW DATABASE SETTING
  'storage.snapshot.interval'` and `SET DATABASE SETTING
  'storage.snapshot.interval'` now return `Unknown setting name` instead of
  `Coordinators don't support snapshots`. Update any tooling that pattern-matched
  on the previous error message.
  [#4066](https://github.com/memgraph/memgraph/pull/4066)
- `schema.node_type_properties()` and `schema.rel_type_properties()` return
  additional columns and changed types. `propertyTypes` is now `List<String>`
  (was untyped), and both procedures include `propertyObservations` and
  `totalObservations` counts. `rel_type_properties()` now partitions results by
  endpoint labels (`sourceNodeLabels`, `targetNodeLabels`), so the same edge
  type connecting different label sets produces separate rows. Clients that
  destructure results by column index or expect the previous column set must be
  updated.
  [#4153](https://github.com/memgraph/memgraph/pull/4153)
- `schema.node_type_properties()` and `schema.rel_type_properties()` renamed
  several `propertyTypes` strings (`Bool` → `Boolean`, `Double` → `Float`,
  `ZonedDateTime` → `DateTime`, `Point2d`/`Point3d` → `Point`), and `mandatory`
  now reflects declared existence constraints instead of sample coverage (always
  `false` for relationships). Update consumers that pattern-matched the old type
  names or treated sampling-based `mandatory` as ground truth.
  [#4186](https://github.com/memgraph/memgraph/pull/4186)
- The JSON metrics endpoint histogram buckets have changed, so p50, p90, and
  p99 latency values may report slightly different numbers compared to 3.10.
  `SHOW METRICS INFO` now returns consistent per-database metrics for the
  currently active database rather than mixing per-database vertex/edge counts
  with aggregate metrics for everything else. Update any alerting thresholds
  that rely on exact histogram values.
  [#3911](https://github.com/memgraph/memgraph/pull/3911)
- The `cross_database` module (and `migrate` aliases) now uses
  `graph.start_timestamp` internally instead of `graph.transaction_id`. The
  `transaction_id` parameter has been removed from the procedure call API. If
  you wrote custom procedures that relied on the `transaction_id` parameter
  passed by the `cross_database` module, switch to `graph.start_timestamp`.
  [#4167](https://github.com/memgraph/memgraph/pull/4167)
- `RELOAD BOLT_SERVER TLS` and the new `RELOAD INTRA_CLUSTER TLS` command now
  require the `RELOAD_TLS` privilege. If you use authorization, grant
  `RELOAD_TLS` to the users or roles that rotate certificates
  (`GRANT RELOAD_TLS TO user;`), otherwise the reload will be rejected.
  [#4154](https://github.com/memgraph/memgraph/pull/4154)
- Session trace no longer writes per-session files under
  `--query-log-directory`. The flag is removed and startup fails if it is still
  set. `SET SESSION TRACE ON` now emits tagged events into the main Memgraph log
  at INFO (when `--log-level` is `info` or lower) instead of a dedicated
  `{uuid}.log` file per session. Remove `--query-log-directory` from configs and
  update tooling that parsed per-session log files to filter the main log by
  `[session=<uuid>]`.
  [#4149](https://github.com/memgraph/memgraph/pull/4149)
- Bare `SHOW STORAGE INFO` now returns only instance-level metrics. Per-database
  fields such as `vertex_count`, `edge_count`, and `storage_mode` moved to
  `SHOW STORAGE INFO ON CURRENT DATABASE` or `SHOW STORAGE INFO ON DATABASE
  <name>`. Several global columns were renamed (`global_memory_tracked` →
  `memory_tracked`, `global_runtime_allocation_limit` → `memory_limit`,
  `global_license_allocation_limit` → `license_memory_limit`, `global_disk_usage`
  → `disk_usage`, `storage_mode` → `global_storage_mode`), and
  `global_isolation_level` / `global_storage_mode` now report startup defaults
  rather than the active database's runtime values. Update scripts and
  dashboards that parsed the bare query to use the per-database variant where
  needed.
  [#4155](https://github.com/memgraph/memgraph/pull/4155)
- `text_search.search`, `search_all`, `search_edges`, and `search_all_edges` no
  longer accept `limit` as a positional argument. Pass `{limit: N}` in the new
  optional `config` map instead (default remains `1000`). Update any queries that
  passed a third positional limit argument.
  [#4191](https://github.com/memgraph/memgraph/pull/4191)
- `IS NULL`, `IS NOT NULL`, `IN`, `STARTS WITH`, `ENDS WITH`, `CONTAINS`, and
  `=~` now bind looser than arithmetic (`*`, `/`, `+`, `-`) and tighter than
  comparisons, matching standard Cypher precedence. Previously these predicates
  bound tighter than arithmetic — for example `(null + 1) * 1 IS NULL` parsed as
  `(null + 1) * (1 IS NULL)` instead of `((null + 1) * 1) IS NULL`. Add
  parentheses if your queries relied on the old binding.
  [#4240](https://github.com/memgraph/memgraph/pull/4240)
- `toBooleanList`, `toFloatList`, and `toIntegerList` now return `null` for
  non-convertible elements instead of failing the whole query. Update queries
  that relied on a bad element aborting the conversion.
  [#4230](https://github.com/memgraph/memgraph/pull/4230)

#### ✨ New features

- Added the `cross_database` module, replacing the deprecated `migrate` module.
  Query other databases (Memgraph, Neo4j, PostgreSQL, MySQL, Oracle, SQL Server,
  S3, Arrow Flight, DuckDB, ServiceNow) directly from Cypher and stream rows
  into your graph. The old `migrate.*` procedure names continue to work as
  aliases.
  [#3832](https://github.com/memgraph/memgraph/pull/3832)
- Added the `derive(path, options)` aggregate function to build an in-query
  virtual graph over matched paths, so graph algorithms can run on derived
  relationships without creating real edges. Supports custom virtual edge types,
  undirected patterns, and property or label overlays on endpoints.
  [#3923](https://github.com/memgraph/memgraph/pull/3923)
- Improved `derive()` so synthetic `sourceNode` and `targetNode` endpoints inherit
  all labels and properties from the original path nodes when those options are
  omitted, making derived graphs easier to build without explicit endpoint config.
  [#4214](https://github.com/memgraph/memgraph/pull/4214)
- Added native OpenMetrics/Prometheus metrics export with per-database
  granularity. Set `--metrics-format=OpenMetrics` to serve metrics at `/metrics`
  in Prometheus scrape format. Each database gets its own query counters,
  latency histograms, and storage gauges labeled by database name. The JSON
  format remains the default but is deprecated.
  [#3911](https://github.com/memgraph/memgraph/pull/3911)
- DEB and RPM packages now ship a stripped production binary plus a separate
  debug-symbol sidecar package instead of distinct `Release` and `RelWithDebInfo`
  variants. Install both when you need debug symbols (equivalent to the previous
  `RelWithDebInfo` install). Docker builds expose the same split via `prod` and
  `debug` flavours.
  [#4079](https://github.com/memgraph/memgraph/pull/4079)
- Vector indexes now support multi-label and wildcard filters. Create indexes on
  any combination of labels with `:L1|L2(prop)` (match any), `:L1&L2(prop)`
  (match all), or `(prop)` (wildcard — every entity with the property). Applies
  to both vertex and edge vector indexes.
  [#3981](https://github.com/memgraph/memgraph/pull/3981)
- `schema.node_type_properties()` and `schema.rel_type_properties()` now return
  richer output including `propertyObservations` and `totalObservations` counts,
  and `propertyTypes` is a proper `List<String>`. `rel_type_properties()`
  additionally partitions results by endpoint labels, so different label-set
  combinations for the same edge type are reported separately.
  [#4153](https://github.com/memgraph/memgraph/pull/4153)
- Added intra-cluster TLS for high availability deployments. New flags
  `--cluster-cert-file`, `--cluster-key-file`, and `--cluster-ca-file` secure
  communication on the management, replication, and coordinator servers. Bolt
  TLS remains independently configured.
  [#4140](https://github.com/memgraph/memgraph/pull/4140)
- Added the `RELOAD INTRA_CLUSTER TLS` Cypher query to rotate intra-cluster TLS
  certificates at runtime without restarting the cluster.
  [#4154](https://github.com/memgraph/memgraph/pull/4154)
- MAIN instances now expose metrics for WAL and snapshot transfer speed,
  including how long REPLICAs take to load durability files, so you can monitor
  replication throughput and spot slowdowns during catch-up or after enabling TLS.
  [#4183](https://github.com/memgraph/memgraph/pull/4183)
- `SHOW TRANSACTIONS` now includes `start_time` (UTC timestamp) and
  `elapsed_ms` columns, making it easier to identify long-running transactions
  without calculating elapsed time manually.
  [#4136](https://github.com/memgraph/memgraph/pull/4136)
- `SHOW TRANSACTIONS` now includes a synthetic `gc` row while storage garbage
  collection is running, with `phase`, `trigger`, `exclusive_lock`, and
  `db_name` metadata, so you can spot a slow or blocking GC directly instead of
  inferring it from latency alone.
  [#4229](https://github.com/memgraph/memgraph/pull/4229)
- Added `SHOW STORAGE INFO ON CURRENT DATABASE` for per-database storage and
  memory metrics without naming the database explicitly. `SHOW STORAGE INFO ON
  DATABASE <name>` (and `ON CURRENT DATABASE`) now work on Community Edition for
  the default database; only non-default databases still require Memgraph
  Enterprise. `SHOW LICENSE INFO` now includes a `memory_limit_policy` row that
  explains how memory limits apply for your license tier.
  [#4155](https://github.com/memgraph/memgraph/pull/4155)
- The OEM license tier is now split into `OEM` (enterprise-equivalent) and
  `OEM_COMMUNITY` (community-tier tracking). Existing OEM license keys continue
  to work and map to the enterprise-equivalent tier on upgrade.
  [#4158](https://github.com/memgraph/memgraph/pull/4158)
- Snapshot creation now logs per-phase progress (edges, vertices, each index
  type, constraints, finalize), the trigger reason, thread count, and final
  statistics. Stalled snapshots can now be diagnosed from the log without
  guessing which phase they are stuck in.
  [#4164](https://github.com/memgraph/memgraph/pull/4164)
- Added fuzzy matching to `text_search.search`, `search_all`, `search_edges`,
  and `search_all_edges` via an optional `config` map (`fuzzy_distance` 0–2,
  `fuzzy_prefix`, `fuzzy_transpositions`), so typo-tolerant and prefix searches
  work without regex. Exact-match behavior is unchanged when fuzzy options are
  omitted.
  [#4191](https://github.com/memgraph/memgraph/pull/4191)
- Session trace events now interleave into the main Memgraph log with
  `[session=]`, `[user=]`, and `[tx=]` tags, so you can follow multiple Bolt
  sessions in one stream without separate log files per session.
  [#4149](https://github.com/memgraph/memgraph/pull/4149)
- Added slow-query and failed-query logging to the main log with `[slow-query]`
  and `[failed-query]` tags. Configure globally with `--log_min_duration_ms`,
  `--log_failed_queries`, and `--log_query_plan`, or override per session via
  `SET SESSION SETTING` for those keys (optional EXPLAIN plan on slow queries).
  [#4211](https://github.com/memgraph/memgraph/pull/4211)
- Added `toBooleanOrNull`, `toFloatOrNull`, `toIntegerOrNull`, and `toStringList`
  conversion functions for safer casting when input may be missing or invalid.
  `toString` and `toStringOrNull` now also stringify spatial points.
  [#4230](https://github.com/memgraph/memgraph/pull/4230)
- Added the `elementId()` function, which returns a node or relationship ID as a
  string for integrations that expect string element IDs (for example Spring
  Data Neo4j). Use `id()` when you need an integer ID and fast point lookups.
  [#4265](https://github.com/memgraph/memgraph/pull/4265)
- The JSON metrics endpoint now has a formally defined contract validated in CI,
  ensuring backward compatibility of metric names and structure across releases.
  [#4143](https://github.com/memgraph/memgraph/pull/4143)

#### 🐞 Bug fixes

- Fixed a bug where `MERGE` followed by `CREATE` in the same query could
  produce unbounded node creation and out-of-memory errors. The underlying scan
  iterator now correctly scopes itself to nodes that existed when the scan
  started.
  [#4080](https://github.com/memgraph/memgraph/pull/4080)
- Fixed `cross_database` (and `migrate`) procedures failing with a `KeyError`
  when used with `USING PERIODIC COMMIT`. The procedures now use a stable
  per-query identifier that does not rotate between batches.
  [#4167](https://github.com/memgraph/memgraph/pull/4167)
- Fixed Memgraph hanging on shutdown when telemetry or license update requests
  were still in flight. Those requests are now aborted during shutdown, and the
  shutdown sequence itself times out after 10 seconds.
  [#4152](https://github.com/memgraph/memgraph/pull/4152)
- Fixed a bug where switching to `IN_MEMORY_ANALYTICAL` storage mode could
  deadlock when async index creation (for example from `ENABLE TTL`) was still
  running. The async indexer now completes before the storage mode lock is taken.
  [#4169](https://github.com/memgraph/memgraph/pull/4169)
- Fixed HA data instances reporting incorrect vertex and storage metrics for the
  default database after joining a cluster. The metrics collector now rebinds the
  default database UUID on join, so per-database counts use the cluster-assigned
  identity.
  [#4172](https://github.com/memgraph/memgraph/pull/4172)
- Fixed a bug where queries with leading whitespace before `EXPLAIN` or
  `PROFILE` could crash or fail to parse. Leading whitespace is now trimmed
  before the keyword is stripped.
  [#4181](https://github.com/memgraph/memgraph/pull/4181)
- Fixed NuRaft coordination logs ignoring `--log-retention-days`. HA coordinator
  log files now rotate and expire using the same retention setting as the main
  Memgraph log, so disk use stays predictable when you tune log retention.
  [#4185](https://github.com/memgraph/memgraph/pull/4185)
- Fixed `schema.node_type_properties()` and `schema.rel_type_properties()`
  reporting property types the Simba BI connector (Power BI, Tableau) could not
  map, which forced columns to VARCHAR. Type strings now match what the connector
  expects so properties surface as Boolean, Float, DateTime, and Point where
  appropriate.
  [#4186](https://github.com/memgraph/memgraph/pull/4186)
- Fixed edge metadata recovery after WAL replay. Edges restored from the WAL are
  now indexed consistently with snapshot-loaded edges, so deletions and id lookups
  with `--storage-enable-edges-metadata` no longer crash or return wrong results.
  [#4189](https://github.com/memgraph/memgraph/pull/4189)
- Fixed HA data instances incorrectly allowing `IN_MEMORY_ANALYTICAL` storage
  mode, which replication does not support. Such instances now fail at startup if
  analytical mode is configured, and runtime `STORAGE MODE IN_MEMORY_ANALYTICAL`
  returns a clear error instead of entering an unsupported state.
  [#4206](https://github.com/memgraph/memgraph/pull/4206)
- Fixed incorrect operator precedence where null checks and string/list
  predicates bound tighter than arithmetic, so mixed expressions such as
  `(null + 1) * 1 IS NULL` could evaluate differently than standard Cypher
  expects.
  [#4240](https://github.com/memgraph/memgraph/pull/4240)
- Fixed startup crashes when the process user does not own the storage directory
  or the configured log file is not writable. Memgraph now exits cleanly with an
  actionable error message — including which user owns the path and how to fix it
  — instead of terminating with a segmentation fault.
  [#4227](https://github.com/memgraph/memgraph/pull/4227)
- Fixed a bug where queries that match nodes with a long chain of OR-ed indexed
  labels (for example `MATCH (n) WHERE n:LabelA OR n:LabelB OR n:LabelC ...`)
  could crash the entire Memgraph process. Planning now stays within the memory
  limit and aborts the query instead of OOM-killing the server.
  [#4246](https://github.com/memgraph/memgraph/pull/4246)
- Fixed a crash that could occur when dropping a database, especially on
  multi-database instances with text indexes.
  [#4188](https://github.com/memgraph/memgraph/pull/4188)
- Fixed a crash when `SHOW STORAGE INFO` computed disk usage while WAL files were
  being rotated or removed. The directory scan now tolerates files disappearing
  mid-scan instead of terminating the server.
  [#4247](https://github.com/memgraph/memgraph/pull/4247)
- Fixed `mgps.await_indexes` adding a one-second delay on every Spark connector
  call. The procedure now returns immediately because index creation is
  synchronous; it remains available for compatibility.
  [#4264](https://github.com/memgraph/memgraph/pull/4264)
- Fixed concurrent `MERGE` ingestion into indexed labels and edge types being
  dramatically slower than equivalent `CREATE` workloads. Parallel upserts now
  run at near-`CREATE` speed instead of degrading when many workers merge
  against the same property index concurrently.
  [#4273](https://github.com/memgraph/memgraph/pull/4273)

### Lab v3.11.0 - June 17th, 2026

### Memgraph v3.10.1 - May 15th, 2026

#### 🐞 Bug fixes

- Fixed Memgraph and MAGE Docker containers failing to start when there are no
  command line arguments provided, e.g. `docker run -it memgraph/memgraph:3.10.0`.
  [#4145](https://github.com/memgraph/memgraph/pull/4145)

### Lab v3.10.2 - May 29th, 2026

### Memgraph v3.10.0 - May 13th, 2026

#### ⚠️ Breaking changes

- Coordinators ignore periodic storage snapshots (no user data to snapshot), and
  `--init-file` / `--init-data-file` are rejected on coordinators and on data
  instances that join an HA cluster (`--management-port` set). Remove those
  flags from coordinator and HA data-instance configs before upgrading, and run
  any bootstrap Cypher against the MAIN after the cluster is formed instead of
  relying on startup init files.
  [#4035](https://github.com/memgraph/memgraph/pull/4035)
- Fine-grained label and edge-type permissions now support explicit `DENY`
  semantics (including `DENY ALL`/`*`) which take precedence over `GRANT`. The
  `UPDATE` permission is split into granular capabilities: `SET LABEL`, `REMOVE
  LABEL`, `SET PROPERTY`, `CREATE EDGE`, `DELETE EDGE`. Existing grants that
  relied on `GRANT NOTHING` / `REVOKE NOTHING` must be migrated to `DENY`.
  [#3789](https://github.com/memgraph/memgraph/pull/3789)
- `SHOW STORAGE INFO` output renames three fields and adds one new field:
  `disk_usage` → `global_disk_usage`, `memory_tracked` → `global_memory_tracked`,
  `allocation_limit` → `global_runtime_allocation_limit`, and adds
  `global_license_allocation_limit`. Update any tooling or scripts that parse
  these field names.
  [#3999](https://github.com/memgraph/memgraph/pull/3999)

#### ✨ New features

- Coordinators now accept read-only system queries (`SHOW CONFIG`, `SHOW LICENSE
  INFO`, `SHOW BUILD INFO`, `SHOW STORAGE INFO`) plus `SET DATABASE SETTING` and
  `RELOAD SSL`, so you can inspect configuration or reload TLS from a
  coordinator without opening a session to a data instance.
  [#4035](https://github.com/memgraph/memgraph/pull/4035)
- Added `SHOW QUERY CALLABLE MAPPINGS` to list all query procedure and function
  aliases with their original source names and types (`procedure` / `function`),
  so you can audit callable mappings without digging through configuration.
  [#4014](https://github.com/memgraph/memgraph/pull/4014)
- Added `WHERE` clause support directly after `YIELD` in procedure calls:
  `CALL mg.procedures() YIELD * WHERE name = 'mg.procedures' RETURN name`. This
  lets you filter procedure results inline instead of wrapping the call in a
  `WITH … WHERE` subquery.
  [#4010](https://github.com/memgraph/memgraph/pull/4010)
- Memgraph now warns (or errors with `--strict-flag-check`) when unexpected
  positional arguments appear after startup flag parsing, catching common
  mistakes like `--flag false` (which silently ignores `false` and sets the flag
  to `true`). Default behavior is strict in development builds; release builds
  emit a warning.
  [#4009](https://github.com/memgraph/memgraph/pull/4009)
- Added `--data-dir-lock-acquisition-timeout-sec` to retry acquiring the
  data-directory file lock for a configurable duration before giving up. This
  prevents spurious startup failures when a pod is force-killed and the old lock
  has not been released yet (e.g. on CephFS or other distributed storage with
  propagation delay). [#4021](https://github.com/memgraph/memgraph/pull/4021)
- Replication error messages now identify exactly which REPLICA failed and why
  (replica name, replication mode, failure reason) instead of generic "at least
  one SYNC/STRICT_SYNC replica has not confirmed" messages, making it faster to
  pinpoint replication issues.
  [#3985](https://github.com/memgraph/memgraph/pull/3985)
- Added `RELOAD BOLT_SERVER TLS` Cypher command to reload SSL/TLS certificates
  at runtime without restarting Memgraph. Existing connections keep the old
  certificate; new connections use the reloaded one. Useful for Let's Encrypt
  renewals and compliance certificate rotations.
  [#3962](https://github.com/memgraph/memgraph/pull/3962)
- `WITH CONFIG` maps (e.g. `CREATE VECTOR INDEX … WITH CONFIG {…}`) now accept
  query parameters as values, so you can pass `{dimension: $dim, capacity: $cap}`
  instead of hard-coding values in the config literal.
  [#3959](https://github.com/memgraph/memgraph/pull/3959)
- `instance_down_timeout_sec` and `instance_health_check_frequency_sec` are now
  Raft-replicated coordinator runtime settings changeable via
  `SET COORDINATOR SETTING` without restarting the cluster. The old
  `--instance-down-timeout-sec` and `--instance-health-check-frequency-sec`
  flags are deprecated and ignored.
  [#3949](https://github.com/memgraph/memgraph/pull/3949)
- HA coordinator cluster operations (register, unregister, promote, demote)
  now follow a Raft-first pattern: the state is committed to the Raft log and
  acknowledged by a majority before returning success. RPCs to data instances
  are retried automatically by the reconciliation loop, so transient network
  issues no longer require manual intervention.
  [#3942](https://github.com/memgraph/memgraph/pull/3942)
  [#3922](https://github.com/memgraph/memgraph/pull/3922)
- Point2D, Point3D, and Enum types are now fully supported in the MGP C, C++,
  and Python query module APIs. Query modules can create, copy, compare, and
  inspect these types as `mgp_value`, fixing crashes in `schema.node_type_properties()`
  on graphs that use spatial or enum properties.
  [#3980](https://github.com/memgraph/memgraph/pull/3980)
- Added `schema.node_type_properties` and `schema.rel_type_properties` config
  argument to match the Neo4j specification, improving compatibility with the
  Neo4j ODBC BI Connector. Also adds the `apoc.version` mapping.
  [#4000](https://github.com/memgraph/memgraph/pull/4000)
- Added server-side descriptions: annotate labels, edge types, properties, and
  databases with human-readable strings that are persisted and visible in
  `SHOW SCHEMA INFO`. Useful for documenting your graph schema directly in the
  database.
  [#3894](https://github.com/memgraph/memgraph/pull/3894)
- Vector edge indexes now use the single-store pattern (storing only a
  `VectorIndexId` in edge properties rather than duplicating vector data),
  matching vertex vector indexes. Snapshot serialization is backward-compatible.
  [#3929](https://github.com/memgraph/memgraph/pull/3929)
- Added GNN import/export module to MAGE: export the graph to PyTorch Geometric
  (PyG) or TensorFlow GNN (TF-GNN) JSON format, run your GNN pipeline
  externally, then write predictions back with `pyg_import` / `tf_import`.
  [#3803](https://github.com/memgraph/memgraph/pull/3803)
- Added Kerberos SSO authentication module: clients can authenticate using
  Kerberos service tickets via GSSAPI, with role mapping through LDAP group
  membership or static principal-to-role configuration.
  [#3916](https://github.com/memgraph/memgraph/pull/3916)
- Added descending (DESC) label-property index support. `ORDER BY … DESC`
  queries can now be served directly by a DESC index, eliminating the sort.
  Create with `CREATE INDEX ON :Label(prop) WITH CONFIG {"order": "DESC"}`.
  [#3996](https://github.com/memgraph/memgraph/pull/3996)
- Added `CALL (<variable-list>) { <subquery> }` syntax for variable-scoped
  subqueries, letting you explicitly control which outer-scope variables are
  accessible inside a `CALL` subquery block.
  [#4073](https://github.com/memgraph/memgraph/pull/4073)
- Added Neo4j-compatible `CREATE INDEX [name] FOR … ON …` and
  `CREATE CONSTRAINT [name] FOR … REQUIRE …` syntax for creating indexes and
  constraints, alongside the existing Memgraph `ON :Label(prop)` / `ASSERT`
  syntax. Both syntaxes are now accepted. Named indexes and constraints emit a
  warning, as names are not yet stored.
  [#4043](https://github.com/memgraph/memgraph/pull/4043)
- Added per-database memory tracking and Tenant Profiles. Every allocation is
  now attributed to its database across three categories — graph storage,
  vector index embeddings, and query execution — visible via
  `SHOW STORAGE INFO ON DATABASE <name>` and `SHOW MEMORY INFO`.
  Create named memory-limit profiles with
  `CREATE TENANT PROFILE "<name>" WITH MEMORY LIMIT <size>`, attach them to
  databases with `SET TENANT PROFILE`, and manage them with `ALTER`, `DROP`,
  `REMOVE TENANT PROFILE FROM DATABASE`, and `SHOW TENANT PROFILES`. Profiles
  are persisted across restarts and replicated to replicas.
  [#3952](https://github.com/memgraph/memgraph/pull/3952)
- Three built-in roles (`admin`, `readonly`, `readwrite`) are now created
  automatically on first enterprise user creation, with the first user assigned
  the `admin` role. Roles can be granted and revoked incrementally using
  `GRANT ROLE[S] <roles> TO <user>` and `REVOKE ROLE[S] <roles> FROM <user>`.
  [#3889](https://github.com/memgraph/memgraph/pull/3889)
- Added compatibility for the Apache Spark Neo4j Connector 5.4.0 by aliasing
  additional query procedures, so PySpark workloads that rely on the connector
  work out of the box.
  [#4091](https://github.com/memgraph/memgraph/pull/4091)
- The embeddings query module can now compute text embeddings via remote APIs
  (OpenAI, Cohere, and other providers supported by `litellm`). Export the
  relevant API key when launching the MAGE container and set `model_name` to
  `"{provider}/{model-name}"` to use a remote model instead of a local one.
  [#4064](https://github.com/memgraph/memgraph/pull/4064)

#### 🛠️ Improvements

- Text index search performance is significantly improved: GIDs are now returned
  directly from the Rust layer instead of full JSON documents, and Tantivy
  searchers are pinned per transaction for snapshot-consistent reads.
  [#3963](https://github.com/memgraph/memgraph/pull/3963)
- Upgraded Tantivy to 0.25, bringing ~30% faster text index imports and ~20%
  better search throughput under contention.
  [#3927](https://github.com/memgraph/memgraph/pull/3927)
- `ORDER BY` is now eliminated when an index scan already provides the required
  ascending order, avoiding an unnecessary sort. With `LIMIT`, this means the
  planner stops after N rows instead of sorting the entire result set first.
  [#3950](https://github.com/memgraph/memgraph/pull/3950)
- Added `AI_PLATFORM` license type that enforces memory limits only on graph
  memory, leaving vector index memory unconstrained up to the system
  `--memory-limit`. Useful when running large vector workloads alongside graph
  data without hitting the overall allocation cap.
  [#3999](https://github.com/memgraph/memgraph/pull/3999)
- Most third-party build dependencies are now managed through Conan 2 packages,
  making builds more reproducible and the contributor setup easier. See the
  updated [build-from-source guide](https://memgraph.com/docs/getting-started/build-memgraph-from-source)
  for changes to the build workflow.
  [#4002](https://github.com/memgraph/memgraph/pull/4002)
- Additional third-party libraries (gflags, NuRaft, Pulsar client, RocksDB,
  usearch) are now managed as Conan recipes, removing the `init` script
  and the `libs/` directory from the build process.
  [#4023](https://github.com/memgraph/memgraph/pull/4023)
- `SHOW INDEX INFO`, `SHOW CONSTRAINT INFO`, `SHOW NODE LABELS`, and
  `SHOW EDGE TYPES` no longer open a storage transaction, reducing lock
  contention and improving responsiveness on busy instances.
  [#4006](https://github.com/memgraph/memgraph/pull/4006)
- Reduced garbage collector contention under high-concurrency supernode
  workloads by moving delta processing to a local copy under a short lock,
  preventing queries from waiting on garbage collection to complete.
  [#4119](https://github.com/memgraph/memgraph/pull/4119)
- Improved garbage collector performance after concurrent import of datasets
  with a significant number of edges on one or more supernodes by avoiding an
  O(n) delta chain walk when unlinking non-sequential deltas.
  [#4126](https://github.com/memgraph/memgraph/pull/4126)

#### 🐞 Bug fixes

- Fixed a potential OOM crash when dropping a large vector index: the internal
  transfer from indexed format back to property store now manages peak memory
  usage so the operation completes cleanly even under tight memory limits.
  [#4001](https://github.com/memgraph/memgraph/pull/4001)
- `RELOAD BOLT_SERVER TLS` is now supported on coordinator instances, so you can
  rotate TLS certificates cluster-wide without opening a session to each data
  instance separately.
  [#4003](https://github.com/memgraph/memgraph/pull/4003)
- Fixed `refactor.merge_nodes` discarding relationship properties: all
  properties on merged relationships are now correctly copied to the new
  relationship.
  [#3948](https://github.com/memgraph/memgraph/pull/3948)
- Fixed `ORDER BY` failing for list values; Cypher list ordering semantics
  (lexicographic / dictionary order) are now applied correctly.
  [#3877](https://github.com/memgraph/memgraph/pull/3877)
- Fixed the `link_prediction` MAGE module returning incorrect ROC-AUC scores
  when all labels belong to a single class (returns a neutral 0.5 AUC instead
  of erroring or producing inconsistent results).
  [#3845](https://github.com/memgraph/memgraph/pull/3845)
- Fixed a use-after-free crash on replicas receiving a snapshot: `Clear()` now
  also discards pending GC deltas, preventing stale pointers from being
  dereferenced on the next transaction commit.
  [#4013](https://github.com/memgraph/memgraph/pull/4013)
- Fixed a crash in HA coordinators caused by a dangling pointer in
  `ReplicationInstanceConnector`; concurrent coordinator operations are now safe
  and cluster membership changes complete without instability.
  [#3922](https://github.com/memgraph/memgraph/pull/3922)
- Fixed a bug where `std::thread::hardware_concurrency()` returning 0 on some
  machines caused RPC and communication servers to be created with zero worker
  threads, leading to silent failures. A safe fallback of 2 is now used.
  [#3982](https://github.com/memgraph/memgraph/pull/3982)
- Fixed replication network channels not being cleanly destroyed during
  shutdown, which could block shutdown or leave sockets open.
  [#4005](https://github.com/memgraph/memgraph/pull/4005)
- Fixed a crash that could occur when calling `SetValueForce` on storage
  failures during a scheduled license revalidation.
  [#4004](https://github.com/memgraph/memgraph/pull/4004)
- Fixed a TCP segmentation bug in SLK multi-file stream processing: when a file
  transition arrives in a fragmented TCP segment, `CheckStreamStatus` now
  validates metadata completeness before signalling `NEW_FILE`, preventing
  incorrect stream state.
  [#3925](https://github.com/memgraph/memgraph/pull/3925)
- Fixed a crash if the telemetry server fails to initialize: the exception is
  now caught, logged, and the instance shuts down gracefully instead of
  terminating abruptly.
  [#4026](https://github.com/memgraph/memgraph/pull/4026)
- Transient disk failures when opening WAL or snapshot output files are now
  handled without crashing; failed snapshots are skipped and retried at the next
  scheduled interval (WAL failures remain fatal).
  [#4025](https://github.com/memgraph/memgraph/pull/4025)
- Fixed log rotation being skipped after instance restarts: spdlog in-memory
  rotation state is now re-initialized on startup so daily log files are
  rotated correctly even after frequent restarts.
  [#4019](https://github.com/memgraph/memgraph/pull/4019)
- Fixed a crash when calling `std::abort` while using an async logger: removing
  the `spdlog::shutdown()` call before abort eliminates a window where the
  logger could segfault, making core dumps more useful for diagnosis.
  [#4034](https://github.com/memgraph/memgraph/pull/4034)
- Fixed a network descriptor leak in the epoll listener and a potential crash
  if `epoll_create1` fails; the latter now throws an exception instead of
  crashing the database.
  [#4032](https://github.com/memgraph/memgraph/pull/4032)
- Fixed fine-grained access control roles leaking across databases: permission
  checks now evaluate only the roles that apply to the currently active
  database, preventing unintended privilege escalation for users with
  per-database roles.
  [#4042](https://github.com/memgraph/memgraph/pull/4042)
- Fixed a replication lag counter overflow that could occur when a REPLICA
  temporarily had more committed transactions than the new MAIN. REPLICAs in
  this transient state are now excluded from the routing table until they
  converge.
  [#4040](https://github.com/memgraph/memgraph/pull/4040)
- Fixed the MAGE DEB package post-install script installing Python packages for
  the wrong architecture on ARM machines.
  [#4093](https://github.com/memgraph/memgraph/pull/4093)
- Fixed index and constraint CREATE/DROP operations not being reverted on abort,
  which could leave ghost entries blocking subsequent retries. This was possible
  only when using STRICT_SYNC replication.
  [#4074](https://github.com/memgraph/memgraph/pull/4074)
- `FREE MEMORY` now reclaims significantly more memory after heavy update/delete
  workloads. Also fixed stale label-property index entries accumulating in
  `IN_MEMORY_ANALYTICAL` mode when repeatedly setting indexed properties.
  [#4071](https://github.com/memgraph/memgraph/pull/4071)
- Fixed a hang or crash when using `LIMIT` on both sides of a `UNION` or
  `UNION ALL` query.
  [#4065](https://github.com/memgraph/memgraph/pull/4065)
- Fixed rare crashes that could occur during shutdown or under heavy load after
  concurrent edge imports.
  [#4020](https://github.com/memgraph/memgraph/pull/4020)
- Fixed a crash or broken memory state caused by jemalloc's lazy thread-local
  storage initialization: if a thread's first allocation triggered an
  allocation-limit check before TLS was ready, jemalloc entered an
  unrecoverable state. Each thread now forces an initial allocation to
  guarantee TLS is set up before limits are enforced.
  [#4095](https://github.com/memgraph/memgraph/pull/4095)
- Hardened the embedded Python runtime against crashes and resource contention
  when running Python query procedures that use libraries like NumPy, PyTorch,
  or DGL. Shutdowns are now more reliable (no more sporadic fatal errors on
  exit) and the dependency scanner no longer pollutes the interpreter's main
  namespace.
  [#4038](https://github.com/memgraph/memgraph/pull/4038)
- Fixed misclassified task priorities that caused `BEGIN`, `COMMIT`, and `ABORT`
  queries to run on reserved high-priority worker threads. Under heavy parallel
  commit/abort load this could starve workers and make the database appear
  blocked to users.
  [#4122](https://github.com/memgraph/memgraph/pull/4122)

### Lab v3.10.1 - May 20th, 2026

### Lab v3.10.0 - May 13th, 2026

### Memgraph v3.9.0

#### ⚠️ Breaking changes

- HA coordinator data directories from previous versions contain corrupted
  cluster configuration log entries that cannot be recovered. When upgrading to
  3.9, all coordinator data directories must be cleaned and coordinators
  restarted fresh; run setup queries again on the new coordinator leader. Data
  instances are not affected. See the [migration
  guide](https://memgraph.com/docs/clustering/high-availability/migrating-to-v3-9-ha).
  [#3900](https://github.com/memgraph/memgraph/pull/3900)

#### ✨ New features

- Vector indexes can now be created with a function in the `WITH CONFIG`
  clause: `CREATE VECTOR INDEX ... WITH CONFIG fun();` so you can generate
  index configuration dynamically instead of using a static map literal.
  [#3801](https://github.com/memgraph/memgraph/pull/3801)
- Added `--storage-floating-point-resolution-bits` to choose double (64), float
  (32), or half (16) precision for floating-point property values system-wide,
  so you can reduce memory use for graphs with many float properties when lower
  precision is acceptable.
  [#3817](https://github.com/memgraph/memgraph/pull/3817)
- Added coordinator setting `deltas_batch_progress_size` to control how many
  deltas a replica processes before sending an in-progress heartbeat during
  replication. HA clusters can tune it to avoid RPC timeouts on large
  transactions (lower value) or reduce overhead on fast replicas (higher
  value). Set via `SET COORDINATOR SETTING`; default is 100000 (unchanged
  behavior for existing clusters).
  [#3797](https://github.com/memgraph/memgraph/pull/3797)
- `SHOW TRANSACTIONS` now lists transactions in committing and aborting phases
  (in addition to running), and shows any snapshot creation in progress with
  phase and progress in metadata, so you can see what is running on the
  database at a glance. [#3833](https://github.com/memgraph/memgraph/pull/3833)
- Added `--logger-type` flag (`sync` or `async`, default `sync`). When set to
  `async`, log messages are written by a background thread instead of inline,
  reducing latency on hot paths under high throughput. Default behavior is
  unchanged; use `--logger-type=async` to opt in.
  [#3844](https://github.com/memgraph/memgraph/pull/3844)
- Added `--log-retention-days` to control how many days daily log files are
  kept before automatic removal. Default is 35 (unchanged from previous
  behavior); set the flag to tune retention for storage or compliance.
  [#3862](https://github.com/memgraph/memgraph/pull/3862)
- `--storage-access-timeout-sec` can now be updated at runtime via the
  `storage.access_timeout_sec` database setting, so you can increase the
  timeout when you hit storage timeout errors without restarting the instance.
  [#3882](https://github.com/memgraph/memgraph/pull/3882)
- Added server-side global parameters: define named values that persist across
  sessions and are resolved in queries. Use `SET GLOBAL PARAMETER`, `UNSET
  GLOBAL PARAMETER`, and `SHOW PARAMETERS` to manage them, and reference values
  in Cypher with `$param` (client-provided parameters take precedence). Global
  parameters are replicated and recovered on restart, so you can use them as a
  durable, cluster-wide store for frequently used query values.
  [#3717](https://github.com/memgraph/memgraph/pull/3717)

#### 🛠️ Improvements

- `SHOW STORAGE INFO` now splits memory into `graph_memory_tracked` and
  `vector_index_memory_tracked`, so you can see how much memory vector indices
  use versus graph data and better plan capacity or troubleshoot memory usage.
  [#3847](https://github.com/memgraph/memgraph/pull/3847)
- Renamed the `SHOW STORAGE INFO` key from `embeddings_memory_tracked` to
  `vector_index_memory_tracked` so the name reflects that it accounts for all
  vector index memory, not only embeddings data.
  [#3863](https://github.com/memgraph/memgraph/pull/3863)
- ARM platforms now install the correct Go and C# binaries for driver tests, so
  driver tests can run on ARM in addition to x86.
  [#3776](https://github.com/memgraph/memgraph/pull/3776)
- Error messages on storage access timeout are now more descriptive so you can
  identify and troubleshoot lock or contention issues more easily.
  [#3778](https://github.com/memgraph/memgraph/pull/3778)
- Reduced memory use for vertices and edges by packing boolean values in the
  delta pointer layout (8 bytes saved per vertex and per edge) with no
  performance impact. [#3814](https://github.com/memgraph/memgraph/pull/3814)
- When `--storage-snapshot-on-exit` is false, any in-progress periodic snapshot
  is aborted on shutdown so the instance stops faster instead of waiting for
  the snapshot to finish. The snapshot digest is now updated only after a
  successful write, so a failed or aborted snapshot no longer causes the next
  periodic snapshot to be skipped.
  [#3823](https://github.com/memgraph/memgraph/pull/3823)
- License selection now considers the database, environment variables, and CLI
  together, automatically choosing the license with the furthest expiry (with
  source priority to break ties), so you no longer need to clear old database
  settings for a newer CLI or env license to apply. Malformed or expired keys
  are rejected immediately when set, and reported license info now matches
  actual feature availability.
  [#3859](https://github.com/memgraph/memgraph/pull/3859)
- EdgeSetProperty WAL delta has been expanded. This speeds up edge recovery and
  replication. Nothing to do, this is all automatic and back-compatible.
  [#3799](https://github.com/memgraph/memgraph/pull/3799)

#### 🐞 Bug fixes

- Improved query planning for BFS with index lookup: a cost model now chooses
  between bidirectional and single-source BFS so you get better execution plans
  for shortest-path queries when both source and destination use an index.
  [#3751](https://github.com/memgraph/memgraph/pull/3751)
- Fixed a bug where `schema.assert` could not be used as a property name or
  value, which blocked backup and restore in some `DUMP DATABASE` scenarios.
  [#3772](https://github.com/memgraph/memgraph/pull/3772)
- When TLS/SSL certificate or key loading fails, Memgraph now logs an error
  message so you can see why startup or connection failed and fix path or
  permission issues. [#3795](https://github.com/memgraph/memgraph/pull/3795)
- With `--data-recovery-on-startup` enabled, the system now recovers both users
  and databases as system metadata on startup, so all system metadata is
  restored consistently; user data is still not recovered.
  [#3807](https://github.com/memgraph/memgraph/pull/3807)
- Plan caching is now enabled again for queries that use BFS, fixing a
  performance regression where a new plan was created for every such query.
  [#3821](https://github.com/memgraph/memgraph/pull/3821)
- User and role details that apply only to enterprise licenses are now
  persisted when switching between enterprise and community, so they are not
  lost and apply again after upgrading back to an enterprise license.
  [#3867](https://github.com/memgraph/memgraph/pull/3867)
- `SHOW SCHEMA INFO` now correctly respects fine-grained label-based access
  control (LBAC) READ permissions, so restricted users see only the schema
  elements they are allowed to access.
  [#3870](https://github.com/memgraph/memgraph/pull/3870)
- Edge removals in the text index are now fully tracked when using `DETACH
  DELETE`, so `SHOW INDEX INFO` reports accurate edge counts instead of
  outdated metadata (search results were already correct).
  [#3878](https://github.com/memgraph/memgraph/pull/3878)
- Concurrent operations on vector indexes (nodes and edges) are now less
  contention-prone: internal locking allows multiple writers in parallel and
  reserves exclusive access only for resize and removal. Garbage collection of
  deleted entries is targeted instead of scanning the full index, improving GC
  performance and fixing a potential crash when vertices or edges were deleted
  in analytical mode before the vector index was cleaned up.
  [#3856](https://github.com/memgraph/memgraph/pull/3856)
- Fixed a bug where the parallel version of `DISTINCT` could invalidate other
  cursors in the query and cause failures. PARALLEL EXECUTION with `DISTINCT`
  now runs correctly and faster.
  [#3876](https://github.com/memgraph/memgraph/pull/3876)
- Fixed inconsistent ordering in the parallel `DISTINCT` cursor, which could
  make PARALLEL EXECUTION output unstable. `DISTINCT` used in PARALLEL
  EXECUTION now produces correct, stable results.
  [#3815](https://github.com/memgraph/memgraph/pull/3815)
- Fixed a bug in HA replication where concurrent edge deltas could be processed
  twice during durability traversal, causing replica inconsistency.
  [#3836](https://github.com/memgraph/memgraph/pull/3836)
- Fixed a crash that could occur when queries using Python user-defined
  functions were cleaned up on internal worker threads. Memgraph now acquires
  the Python GIL before releasing Python object references during query
  teardown, preventing segfaults when using Python query modules with custom
  functions. [#3837](https://github.com/memgraph/memgraph/pull/3837)
- Fixed a rare race in durability when a transaction that created concurrent
  edges aborted while the delta chain was being processed, which could cause
  duplicate or missed edge-delta encoding and durability integrity issues.
  [#3839](https://github.com/memgraph/memgraph/pull/3839)
- Fixed a segfault that could occur when closing Python query modules during
  Memgraph shutdown. A missing Python interpreter state check is now performed
  before cleanup so shutdown completes cleanly when Python modules are loaded.
  [#3842](https://github.com/memgraph/memgraph/pull/3842)
- Fixed a segfault during vector index garbage collection when vertices or
  edges were being inserted concurrently, which could occur with bulk
  create/delete/create workloads.
  [#3846](https://github.com/memgraph/memgraph/pull/3846)
- Fixed a data race in the high-availability coordinator's cluster
  configuration handling. Concurrent config loads and saves are now properly
  synchronized, preventing inconsistent state and coordinator instability
  during cluster membership changes.
  [#3850](https://github.com/memgraph/memgraph/pull/3850)
- Shutdown on SIGTERM and SIGINT no longer runs inside a signal handler, where
  logging and mutex use were unsafe. Signals are now consumed on the main
  thread so the full shutdown sequence runs in normal context, making shutdown
  more robust against rare crashes or hangs.
  [#3851](https://github.com/memgraph/memgraph/pull/3851)
- Fixed a Python binding assertion caused by unhandled import errors. When the
  system encounters errors (for example when over the memory limit), Python will
  fail to load modules. This was not gracefully handled and would cause
  termination. The Python bindings are now more robust.
  [#3864](https://github.com/memgraph/memgraph/pull/3864)
- Fixed a bug where periodic commit failure caused a double abort (which in
  turn caused an assert). Periodic commit failure is a special case and is now
  handled correctly by the interpreter. Users can run periodic commit without
  worrying about query failures.
  [#3868](https://github.com/memgraph/memgraph/pull/3868)
- Fixed a bug where incorrect queries using CALL + UNION would not throw an
  exception but would just produce wrong results.
  [#3869](https://github.com/memgraph/memgraph/pull/3869)
- Fixed occasional "Operation not permitted" error when vertices with
  concurrently imported edges are indexed.
  [#3871](https://github.com/memgraph/memgraph/pull/3871)
- Communication data parsing is now more robust when values do not match the
  expected type, so you get fewer unexpected communication errors between
  clients and the database.
  [#3873](https://github.com/memgraph/memgraph/pull/3873)
- The schema query module now works with ZonedDateTime properties, so you can
  use schema assertions and schema-based workflows (such as backup and restore)
  on graphs that store temporal data with time zones.
  [#3874](https://github.com/memgraph/memgraph/pull/3874)
- Fixed a bug where HA coordinator Raft cluster configuration log entries were
  not persisted correctly, causing coordinators to crash-loop on restart if
  uncompacted configuration entries existed on disk — most commonly triggered
  when a newly added coordinator restarts before a Raft snapshot compacts those
  entries. Users running HA clusters who experience a coordinator crash-loop
  after adding a new coordinator should upgrade to this version; all
  coordinator data directories should be wiped before restarting and setup
  queries should be run again on the new coordinator leader.
  [#3900](https://github.com/memgraph/memgraph/pull/3900)
- Fixed a crash in the `embeddings` query module when multiple Bolt sessions
  called `embeddings.text()` or `embeddings.node_sentence()` in parallel. Model
  initialization is now thread-safe with a global cache using double-checked
  locking, preventing race conditions and eliminating redundant model
  instantiation per request.
  [#3914](https://github.com/memgraph/memgraph/pull/3914)
- Python module cleanup during shutdown no longer crashes the process;
  previously, filesystem operations that failed while removing `__pycache__`
  directories would throw exceptions inside a destructor, triggering
  `std::terminate` and an abrupt abort — now those operations use non-throwing
  error-code overloads and log warnings instead, so shutdown completes
  gracefully even when cached files have already been removed.
  [#3908](https://github.com/memgraph/memgraph/pull/3908)
- Fixed a crash that could occur when a coordinator's internal Raft consensus
  thread encountered an error during startup or operation — the process would
  terminate with a segmentation fault instead of exiting cleanly. This affected
  HA deployments where a coordinator node hit an unexpected state in its
  background commit path. No user action is required; the fix ensures the
  coordinator process exits immediately and cleanly without triggering a
  cascading failure in global cleanup.
  [#3904](https://github.com/memgraph/memgraph/pull/3904)
- Fixed a race condition during graceful shutdown (SIGTERM) of
  high-availability data instances where the internal management RPC server
  could process incoming coordinator health-check requests after the
  replication state had already been destroyed, resulting in a crash. The
  shutdown sequence now correctly stops the management server before tearing
  down replication state, so users running HA clusters should no longer observe
  unexpected crashes or error logs when stopping data instances with SIGTERM.
  No configuration changes or user action is required.
  [#3903](https://github.com/memgraph/memgraph/pull/3903)
- Fixed a bug in vector index garbage collection where the `equal_range`
  iterator incorrectly yielded tombstone (deleted) entries as live matches,
  causing double-frees that corrupted size accounting and led to crashes. Users
  running workloads with frequent deletions on vector indexes were at risk of
  hitting this crash. [#3899](https://github.com/memgraph/memgraph/pull/3899)

### Lab v3.9.0 - March 18th, 2026

### Memgraph v3.8.1 - February 17th, 2026

#### 🐞 Bug fixes

- Fixed using `LOAD CSV` via SSL, by adding missing `ca-certificates` package
  to the Memgraph Docker image
  [#3793](https://github.com/memgraph/memgraph/pull/3793)
- A data instance could get into a deadlock state when some specific timings
  are triggered. This is now fixed so users should be able to always start a
  data instance without any deadlocks occurring.
  [#3787](https://github.com/memgraph/memgraph/pull/3787)
- Fix so that concurrent edge imports will no longer delay garbage collection.
  This resulted in temporary memory spikes after import, until the periodic
  garbage collector could free up memory.
  [#3784](https://github.com/memgraph/memgraph/pull/3784)

### Memgraph v3.8.0 - February 11th, 2026

#### ⚠️ Breaking changes

- Switching to analytical mode now fails if constraints exist. Creating
constraints or text indexes in analytical mode is now blocked. Workflows that
relied on these operations will need to be updated.
[#3461](https://github.com/memgraph/memgraph/pull/3461)
- Removed deprecated flags `--auth-module-executable`,
`--bolt_session_inactivity_timeout`, and `--storage_parallel_index_recovery`.
Users should migrate to `--auth-module-mappings` and
`--storage-parallel-schema-recovery` respectively.
[#3474](https://github.com/memgraph/memgraph/pull/3474)
- The auth module fix introduced the new message ID protocol. All the external
auth modules have to be updated to support the new protocol.
[#3755](https://github.com/memgraph/memgraph/pull/3755)

#### ✨ New features

- Concurrent edge creation on supernodes is no longer treated as a serialization
error, allowing truly concurrent edge-heavy batch imports and improving import
speeds for high-degree vertices by eliminating transaction aborts and retries.
[#3103](https://github.com/memgraph/memgraph/pull/3103)
- Added `toFloatList()`, `toBooleanList()`, and `toIntegerList()` functions for
converting list elements to the specified types.
[#3318](https://github.com/memgraph/memgraph/pull/3318)
- Introduced parallel Cypher execution as an Enterprise feature. Queries can opt
in via `USING PARALLEL EXECUTION` (with an optional thread count) to run scan
and aggregation phases on multiple worker threads for faster execution of large
queries. Requires the `PARALLEL_EXECUTION` privilege.
[#3394](https://github.com/memgraph/memgraph/pull/3394)
- Simplified data import processes by introducing the native `LOAD JSONL`
clause, enabling users to ingest JSON Lines files directly into the database.
This eliminates the need for external preprocessing or custom transformation
scripts, significantly reducing the time and complexity required to import
large, line-delimited datasets.
[#3410](https://github.com/memgraph/memgraph/pull/3410)
- Parquet files can now be downloaded from any `http(s)/ftp` server directly
when using `LOAD PARQUET` clause.
[#3444](https://github.com/memgraph/memgraph/pull/3444)
- Streamlined data ingestion workflows by allowing the direct loading of JSONL
files from HTTP and FTP servers. This update includes a new
`file.download_conn_timeout_sec` run-time setting, giving users greater control
over connection stability and preventing hung processes during remote data
transfers. [#3453](https://github.com/memgraph/memgraph/pull/3453)
- Added support for loading JSONL and CSV files directly from S3. Users can now
load graphs from more sources. Furthermore, these features improve the so-called
on-demand capabilities of Memgraph.
[#3463](https://github.com/memgraph/memgraph/pull/3463)
- Extended pattern comprehension support to work correctly in all query clauses
(`RETURN`, `WITH`, `WHERE`, `ORDER BY`, `UNWIND`, `FOREACH`, `CREATE`, `SET`,
`MERGE`) and added support for nested pattern comprehensions.
[#3495](https://github.com/memgraph/memgraph/pull/3495)
- Simplified High Availability (HA) deployments by introducing the `UPDATE
CONFIG FOR` command. By allowing the leader coordinator to modify Bolt server
settings across instances, this feature enables a one-step installation process
for HA Helm charts. [#3530](https://github.com/memgraph/memgraph/pull/3530)
- Enhanced visibility into session identity and access control by introducing
the `username()` and `roles()` built-in functions. These functions allow users
to programmatically retrieve current authentication details, simplifying
auditing and the development of dynamic, role-based logic within queries.
[#3563](https://github.com/memgraph/memgraph/pull/3563)
- Added `--storage-backup-dir-enabled` flag to control whether the `.old`
directory is used for storing old durability files during potential data loss
scenarios. When disabled, old durability files are deleted instead of backed up.
[#3631](https://github.com/memgraph/memgraph/pull/3631)
- Temporal functions (`date`, `duration`, `localdatetime`, `datetime`, and
`localtime`) now accept null as an argument, returning null instead of throwing
an error. [#3714](https://github.com/memgraph/memgraph/pull/3714)
- Added `load_from_str()` function to the `json_util` query module, allowing
JSON to be loaded directly from a string instead of a file.
[mage#530](https://github.com/memgraph/mage/pull/530)
- Added an LLM query module with a completion function that uses LiteLLM to send
text to an LLM and receive a completion. This enables LLM-powered summarization
or text generation directly from Cypher; users configure their provider API key
and optionally the model or API endpoint before calling the function.
[#3761](https://github.com/memgraph/memgraph/pull/3761)

#### 🛠️ Improvements

- Optimized vector property storage by eliminating data duplication between the
property store and vector index. Vector data is now stored only once in the
index, significantly reducing memory usage for vector-indexed properties.
[#3291](https://github.com/memgraph/memgraph/pull/3291)
- Updated telemetry to include anonymous data for role counts and label/node
distributions. [#3424](https://github.com/memgraph/memgraph/pull/3424)
- Added a caching mechanism for regexes. Queries that use regexes inside the
where clause (=~ operator) should be significantly faster.
[#3468](https://github.com/memgraph/memgraph/pull/3468)
- Reduced size of PropertyValue, meaning less memory is used during a query
[#3473](https://github.com/memgraph/memgraph/pull/3473)
- Changed dependency manager used in GraphQL tests to `pnpm`, which is hardened
against any supply-chain vulnerabilities in the npm ecosystem.
[#3501](https://github.com/memgraph/memgraph/pull/3501)
- Strengthened database security by adding a `SECURITY` clause to triggers,
allowing administrators to define whether a trigger executes with the
permissions of the user who created it (`DEFINER`) or the user who triggered the
event (`INVOKER`). This provides granular control over access rights and
prevents unauthorized data modifications during automated workflows.
[#3503](https://github.com/memgraph/memgraph/pull/3503)
- Merged MAGE and mgcxx repos into the memgraph repo. From the usage or
packaging perspective, everything stayed the same. From the internal development
perspective, it should be easier to develop memgraph because everything is
centralized now. [#3574](https://github.com/memgraph/memgraph/pull/3574)
- Boosted performance during snapshot and WAL file processing by optimizing
buffer reuse. By eliminating unnecessary buffer discards when data is already
within the loaded range, users will experience more efficient file loading and
faster system recovery times.
[#3591](https://github.com/memgraph/memgraph/pull/3591)
- Simplified disaster recovery and cloud integration by enabling the `RECOVER
SNAPSHOT` query to pull data directly from S3, HTTP, HTTPS, or FTP sources. This
eliminates the need for manual staging of snapshot files and supports flexible
AWS authentication via config maps, environment variables, or run-time settings.
[#3596](https://github.com/memgraph/memgraph/pull/3596)
- Added two new flags to control RocksDB diagnostic:
`--storage-rocksdb-info-log-level` and
`--storage-rocksdb-enable-thread-tracking`. It's possible to reduce the
gettimeofday/gettid syscalls overhad from RocksDB making Memgraph faster in
cases when that overhead is visible under your workload.
[#3609](https://github.com/memgraph/memgraph/pull/3609)
- Improved system availability during schema modifications by allowing read
operations to run concurrently with constraint creation and removal. This change
replaces the requirement for UNIQUE access with READ_ONLY access, preventing
read queries from being blocked during these operations.
[#3615](https://github.com/memgraph/memgraph/pull/3615)
- Improved concurrency for snapshot creation by switching from WRITE to READ
access. Snapshots no longer block READ_ONLY operations (e.g., creating
constraints), allowing them to execute simultaneously.
[#3625](https://github.com/memgraph/memgraph/pull/3625)
- Integrated [mgcxx](https://github.com/memgraph/mgcxx) into the Memgraph repo
to simplify development and reduce build complexity.
[#3636](https://github.com/memgraph/memgraph/pull/3636)
- Upgraded cuGraph integration from RAPIDS 22.02 / CUDA 11.5 to RAPIDS 25.12 /
CUDA 13.1, bringing GPU-accelerated graph algorithms up to date with current
NVIDIA tooling. [#3709](https://github.com/memgraph/memgraph/pull/3709)
- Added cuGraph MAGE images to regular builds. Users can now pull
`memgraph/memgraph-mage:<version>-relwithdebinfo-cugraph` and run with GPU
access for GPU-accelerated graph algorithms. Requires NVIDIA GPU with driver
`>=590` and CUDA toolkit `>=13.1`.
[#3723](https://github.com/memgraph/memgraph/pull/3723)
- Fixed performance regression in all-shortest paths queries by avoiding
expensive exception handling.
[#3746](https://github.com/memgraph/memgraph/pull/3746)
- Fixed performance regression for vector index creation and import by disabling
SimSIMD for the vector index metric and simplifying index storage to a single
map keyed by index ID, so vector index build and import are faster again while
keeping SIMD benefits. [#3754](https://github.com/memgraph/memgraph/pull/3754)
- Fixed spelling errors across MAGE codebase and documentation.
[mage#227](https://github.com/memgraph/mage/pull/227)
- Improved error messages for `LOAD PARQUET`, `LOAD JSONL`, `LOAD CSV`, and
remote snapshot loading so users can diagnose load failures more easily.
[#3758](https://github.com/memgraph/memgraph/pull/3758)
- Memgraph packages now include the latest `mgconsole` binary. See [mgconsole
v1.5.0](https://github.com/memgraph/mgconsole/releases/tag/v1.5.0) for
improvements.  [#3764](https://github.com/memgraph/memgraph/pull/3764)

#### 🐞 Bug fixes

- Comparisons of incompatible types (e.g., string vs number) now return null
instead of raising an error, allowing incompatible comparisons to be captured
and handled in Cypher rather than causing queries to abort.
[#2974](https://github.com/memgraph/memgraph/pull/2974)
- Fixed replica recovery using an incorrect snapshot timestamp, which could
cause WAL files to be skipped during the recovery process. The system now
correctly uses the durable timestamp instead of the start timestamp when
creating recovery steps for replicas.
[#3431](https://github.com/memgraph/memgraph/pull/3431)
- Analytical mode now enforces stricter safety rules. Users can no longer switch
to analytical mode if constraints exist, create constraints or text indexes in
analytical mode, and a snapshot is automatically created when switching back to
transactional mode. These changes prevent unsafe or corrupted states caused by
unsupported analytical-mode operations.
[#3461](https://github.com/memgraph/memgraph/pull/3461)
- Fixed bug where `CALL { RETURN 1 AS x } RETURN *;` returned an empty set
instead of the subquery result. `RETURN *` now correctly includes variables from
`CALL` subqueries. [#3481](https://github.com/memgraph/memgraph/pull/3481)
- Removed manual clean-up of composite and nested indices in end-to-end tests.
Manual deletion of such indices are no longer needed, since`GQLAlchemy` now
handles these automatically.
[#3482](https://github.com/memgraph/memgraph/pull/3482)
- Fixed data races in vertex edge list access, gatekeeper notification timing,
and signal handler by adding proper locking and removing non-async-signal-safe
operations. [#3484](https://github.com/memgraph/memgraph/pull/3484)
- Fixed non-repeatable reads in strict sync replication mode. Transactions on
replicas now correctly maintain read consistency during two-phase commit.
[#3504](https://github.com/memgraph/memgraph/pull/3504)
- Fixed various unlikely crashes related to concurrency and thread-safety.
[#3509](https://github.com/memgraph/memgraph/pull/3509)
- Accessing properties of a deleted node/relationship (e.g., `MATCH (n) DETACH
DELETE n RETURN n {.*};`) now raises an error instead of returning invalid data.
If you need properties from deleted objects, capture them before deleting.
[#3522](https://github.com/memgraph/memgraph/pull/3522)
- Fixed triggers not checking label-based access control. Users should expect
appropriate access controls under triggers.
[#3569](https://github.com/memgraph/memgraph/pull/3569)
- Increased cluster stability during role transitions by ensuring smooth
demotion of the main instance (even during a long-running write transaction).
The system now waits for all active write transactions to finish before
finalizing a demotion, preventing GID (Global ID) overlaps and ensuring data
consistency even when handling long-running transactions.
[#3594](https://github.com/memgraph/memgraph/pull/3594)
- Reduced memory overhead on strict sync replicas by ensuring transaction deltas
are released immediately upon completion. This optimization prevents temporary
memory spikes and ensures more stable, predictable resource usage by bypassing
the delay of the standard garbage collection cycle.
[#3595](https://github.com/memgraph/memgraph/pull/3595)
- Fixed potential crash caused by recursive malloc during memory allocation
hooks. TLS allocation is now forced at startup to prevent this issue.
[#3601](https://github.com/memgraph/memgraph/pull/3601)
- Optimized disk space usage by ensuring aborted snapshots are deleted
immediately upon failure. This eliminates storage "pollution" and provides more
efficient, predictable disk management.
[#3602](https://github.com/memgraph/memgraph/pull/3602)
- Fixed potential memory leak during network resolution failure. Users can
expect that the number of opened files won't grow even in extreme cases when
Memgraph reports system errors during network resolution.
[#3620](https://github.com/memgraph/memgraph/pull/3620)
- Enhanced storage efficiency in HA environments by automatically clearing
`tmp` directories on restart and deleting WAL files immediately after request
processing. This ensures more predictable disk usage and prevents storage
accumulation from temporary files.
[#3621](https://github.com/memgraph/memgraph/pull/3621)
- Fixed a deadlock issue in strict sync replication where the cached accessor
used in two-phase commit was not properly destroyed when a replica gets
promoted. This prevents deadlocks in scenarios where a replica was demoted,
promoted, and demoted again before the previous commit finished.
[#3628](https://github.com/memgraph/memgraph/pull/3628)
- Fixed race condition when adding log sinks by using thread-safe
`dist_sink_mt`.  Logging is also more efficient with a single mutex instead of
three.  [#3634](https://github.com/memgraph/memgraph/pull/3634)
- Fixed text index cleanup to correctly release all Tantivy resources before
removing index directories. Deleting text indexes no longer fails.
[#3641](https://github.com/memgraph/memgraph/pull/3641)
- Fixed crash when removing properties on type-constrained vertices. Setting a
property to NULL (i.e., removing it) on a vertex with a type constraint no
longer causes an assertion failure. Type constraints enforce the property's type
only when a value is present, they don't require the property to exist.
[#3644](https://github.com/memgraph/memgraph/pull/3644)
- Fixed crash when renaming a non-existing database to its current name.
Renaming a database to the same name now throws an exception instead.
[#3647](https://github.com/memgraph/memgraph/pull/3647)
- Fixed crash caused by corrupted expressions during query planning. Queries
with `OPTIONAL MATCH`, `CALL` subqueries, and `WHERE` clauses containing `AND`
with `id()` no longer crash the server.
[#3649](https://github.com/memgraph/memgraph/pull/3649)
- Fixed an issue where the `migrate` module would fail when returning more than
1000 results. [#3706](https://github.com/memgraph/memgraph/pull/3706),
[mage#711](https://github.com/memgraph/mage/pull/711)
- Fixed `path.subgraph_all` to properly expand beyond the start node when
`filterStartNode == false`, even if the start node doesn't meet `labelFilter`
criteria. Negative values for `maxHops` are now ignored.
[#3707](https://github.com/memgraph/memgraph/pull/3707),
[mage#613](https://github.com/memgraph/mage/pull/613)
- Fixed an issue where snapshot files could be unintentionally deleted during
normal snapshot creation due to an incorrect abort check. This could result in a
“Snapshot creation successful” message even though no snapshot file was actually
written to disk. Snapshot creation is now reliable and only deletes files when
an abort is explicitly requested.
[#3711](https://github.com/memgraph/memgraph/pull/3711)
- Improved index selection to prefer indices that satisfy more query filters,
reducing post-filtering overhead. Composite indices are now correctly chosen
when they provide better selectivity.
[#3712](https://github.com/memgraph/memgraph/pull/3712)
- Fixed segfault on shutdown when running many databases (500+). Global
replication clients are now properly shut down before databases are destroyed.
[#3713](https://github.com/memgraph/memgraph/pull/3713)
- Improved snapshot and WAL retention policy. WAL files with incorrect UUIDs are
now deleted, and corrupted or incomplete snapshots are treated like any other
snapshot, keeping only the latest `--storage-snapshot-retention-count` files
regardless of validity. [#3718](https://github.com/memgraph/memgraph/pull/3718)
- Fixed coordinator crash when adding itself as a coordinator instance fails.
The system now throws an exception instead of terminating, preserving the valid
coordinator state. [#3727](https://github.com/memgraph/memgraph/pull/3727)
- Fixed `--also-log-to-stderr` flag not being respected during Memgraph
initialization. Users can now see startup logs and crash messages in the
terminal when this flag is enabled.
[#3728](https://github.com/memgraph/memgraph/pull/3728)
- Enterprise license key and organisation name set through environment variables
or command-line flags are now visible in `SHOW DATABASE SETTINGS` output. This
helps users verify their license configuration is correctly applied.
[#3730](https://github.com/memgraph/memgraph/pull/3730)
- Fixed a potential crash when accessing server endpoint before the server was
fully initialized. [#3743](https://github.com/memgraph/memgraph/pull/3743)
- Fixed data instance registration when a main instance already exists. The
coordinator now correctly sends registration information to new data instances.
[#3744](https://github.com/memgraph/memgraph/pull/3744)
- Aligned Bolt exception handling across all query paths. Replication failures
now always surface as `ClientError` instead of varying between `ClientError` and
`TransientError` depending on transaction mode, so clients can handle
replication errors consistently.
[#3747](https://github.com/memgraph/memgraph/pull/3747)
- Stale responses from auth modules are now ignored. Previously, stale data
could be read from the module depending on communication errors, potentially
leading to logging in as the wrong user. External auth modules must be updated
to support the new message ID protocol.
[#3755](https://github.com/memgraph/memgraph/pull/3755)
- Fixed boolean arguments being negated when executing Rust query module
procedures. [mage#447](https://github.com/memgraph/mage/pull/447)
- `toFloat` and `toFloatList` now accept Boolean values, converting `true` to
`1.0` and `false` to `0.0`, so expressions and pipelines that expect floats no
longer fail when given Booleans.
[#3759](https://github.com/memgraph/memgraph/pull/3759)
- The MAGE Python embeddings module now correctly handles `excluded_properties`
when Cypher passes it as a tuple. Users can rely on property exclusion when
generating embeddings without encountering runtime errors.
[#3763](https://github.com/memgraph/memgraph/pull/3763)
- Fixed deadlock when using `LIMIT` in queries that perform `LOAD PARQUET` or
other load operations. Load queries with `LIMIT` now complete without hanging.
[#3758](https://github.com/memgraph/memgraph/pull/3758)
- Fixed `VectorIndexId` to `mgp_value` type conversion so query modules and
procedures that work with vector indexes receive and return correct values
instead of incorrect or failing conversions.
[#3767](https://github.com/memgraph/memgraph/pull/3767)
- Fixed `roles()` and `username()` returning wrong values when used inside
triggers with `SECURITY DEFINER`. Triggers now report the definer's identity
correctly for auditing and role-based logic.
[#3768](https://github.com/memgraph/memgraph/pull/3768)

### Lab v3.8.0 - February 11th, 2026

### Memgraph v3.7.2 - December 23rd, 2025

#### 🐞 Bug fixes

- Vector index recovery can now run in parallel, with additional improvements to
  index resizing and safer entry removal logic. This results in faster startup
  and schema recovery for databases with large vector indexes. Users can enable
  this behavior by starting Memgraph with
  `--storage-parallel-schema-recovery=true`.
  [#3562](https://github.com/memgraph/memgraph/pull/3562)
- Complete shutdown of coordinators and Settings held locally under main().
  Avoids segfaults while destroying static objects in an uncontrolled order.
  Avoids unexpected errors from nuraft while shutting down or exiting.
  [#3567](https://github.com/memgraph/memgraph/pull/3567)
- Fixed a bug where using SSO modules would cause username+password users to
  have wrong permissions.
  [#3492](https://github.com/memgraph/memgraph/pull/3492)
- Text index recovery now works correctly from snapshots even when the index
  directories are missing, allowing users to restore full-text indexes using
  only a snapshot, which was not possible before. Additionally, the Clear()
  function used during DROP GRAPH has been made safer to prevent intermediate
  system states in case of failures.
  [#3570](https://github.com/memgraph/memgraph/pull/3570)
- Snapshots can now be created on replicas too manually by user or by the
  scheduler thread. Users can expect a faster recovery on replica from a
  snapshot compared to a recovery from sequential WAL files only.
  [#3420](https://github.com/memgraph/memgraph/pull/3420)
- Fixed an AddressSanitizer-detected memory leak in HTTP POST requests by
  properly freeing the curl headers list.
  [#3586](https://github.com/memgraph/memgraph/pull/3586)

### MAGE v3.7.2 - December 23rd, 2025

- Version bump to v3.7.2 to match the latest Memgraph version.

### Memgraph v3.7.1 - November 25th, 2025

#### ⚠️ Breaking changes

- Changed: Uses documented grammar for fine-grained edge permissions, such as
  `GRANT READ ON EDGES OF TYPE :FOLLOWS TO alice` or
  `GRANT UPDATE ON EDGES OF TYPE * TO bob`. Due to the new syntax, this is a
  breaking change.[#3457](https://github.com/memgraph/memgraph/pull/3457)

#### 🛠️ Improvements

- Forced the GC to acquire a shared lock in analytical mode, preventing
  unnecessary blocking of other transactions and improving responsiveness.
  A new runtime flag, `storage-gc-aggressive`, allows users to opt into
  aggressive GC, which performs full cleanup of unused objects but requires
  taking a unique lock that will temporarily block the system.
  [#3448](https://github.com/memgraph/memgraph/pull/3448)
- Edge creation is refactored to ensure skip list GC doesn't run while vertex
  locks are held. This results in better throughput and fewer stalls under heavy
  concurrency. No user action is required to benefit from the improved
  performance. [#3441](https://github.com/memgraph/memgraph/pull/3441)
- Improved skip list GC performance by optimizing how deleted objects are
  tracked. This results in faster cleanup and reduced overhead during intensive
  write workloads. [#3421](https://github.com/memgraph/memgraph/pull/3421)

### MAGE v3.7.1 - November 25th, 2025

- Version bump to v3.7.1 to match the latest Memgraph version.

### Lab v3.7.1 - December 19th, 2025

### Lab v3.7.0 - December 5th, 2025

### Memgraph v3.7.0 - November 19th, 2025

#### ⚠️ Breaking changes

- IMPORTANT: Changed fine-grained access configuration syntax and semantics to be more
  flexible. Permissions can be set on groups of labels matching any or exactly
  what is set in the `GRANT` statement. `CREATE`, `READ`, `UPDATE`, and
  `DELETE` permissions are no longer organised in a hierarchy and can be
  granted or revoked independently. This is a breaking change, and the
  fine-grained access syntax and semantics of Memgraph 3.6 and earlier are no
  longer supported. Migrated databases must be manually configured to grant
  permissions based on the new rules. To migrate, please follow the [MIGRATION
  GUIDE](https://memgraph.com/docs/database-management/authentication-and-authorization/mlbac-migration-guide).
  [#3343](https://github.com/memgraph/memgraph/pull/3343)
- Changed the way Memgraph is built (only relevant if you are building Memgraph
  from source) and dropped support for Debian 11.
  [#2691](https://github.com/memgraph/memgraph/pull/2691)

#### ✨ New features

- Enable InList optimisation (caching a set for O(1) lookup) on more
  expressions. Now `x IN range(0,1000)` and other expressions benefit from this
  optimisation. [#3347](https://github.com/memgraph/memgraph/pull/3347)
- Added support for loading Parquet files directly from the local filesystem or
  S3-like storage. Users can now use `LOAD PARQUET FROM /path/to/file.parquet;`
  in their Cypher queries for efficient bulk data import. The feature enhances
  data loading options, eliminates the need for format conversion, and provides
  significantly faster import performance for large datasets.
  [#3356](https://github.com/memgraph/memgraph/pull/3356)
  [#3378](https://github.com/memgraph/memgraph/pull/3378)
  [#3395](https://github.com/memgraph/memgraph/pull/3395)
- Allow the use of a self-signed identity server when authenticating via custom
  SSO. Users can now add their identity server's certificates via the
  `MEMGRAPH_SSO_CUSTOM_OIDC_EXTRA_CA_CERTS` env variable.
  [#3389](https://github.com/memgraph/memgraph/pull/3389)

#### 🛠️ Improvements

- Added new toolchain (v7) with OpenMP, updated `clang`, `cmake`, `boost` and
  more. Updated Linux distributions support: added Fedora 42 (x86_64 and
  aarch64), Debian 13 (x86_64 and aarch64) and Rocky 10 (x86_64 only); dropped
  support for Debian 11.
  [#2691](https://github.com/memgraph/memgraph/pull/2691)
- Expanded the list of supported Linux distributions by modifying bash scripts
  to allow building on distributions that are binary compatible with supported
  ones (e.g., Linux Mint -> Ubuntu; AlmaLinux -> CentOS). Fixed the IPO not
  supported issue. [#3231](https://github.com/memgraph/memgraph/pull/3231)
- Introduced more precise and efficient internal list representations instead
  of storing all elements as generic PropertyValue lists. As a result,
  operations that create deltas for numeric lists will consume less memory,
  improving overall efficiency. In some cases, such as lists of small integers
  or short non-numeric lists, memory usage may increase slightly due to
  additional metadata or the fixed 4-byte integer representation.
  [#3282](https://github.com/memgraph/memgraph/pull/3282)
- Changed fine-grained access configuration syntax and semantics to be more
  flexible. Permissions can be set on groups of labels matching any or exactly
  what is set in the `GRANT` statement. `CREATE`, `READ`, `UPDATE`, and
  `DELETE` permissions are no longer organised in a hierarchy and can be
  granted or revoked independently. This is a breaking change, and the
  fine-grained access syntax and semantics of Memgraph 3.6 and earlier are no
  longer supported. Migrated databases must be manually configured to grant
  permissions based on the new rules.
  [#3343](https://github.com/memgraph/memgraph/pull/3343)
- Added `relwithdebinfo` builds which use `malloc` instead of `jemalloc`; also
  included `heaptrack` in all `relwithdebinfo` Docker images.
  [#3346](https://github.com/memgraph/memgraph/pull/3346)
- Improved `RECOVER SNAPSHOT` by not backing up snapshot and WAL files
  indefinitely; instead, a single backup is performed. There will now be a
  single .old directory under the snapshots and wals directories. Files already
  present in the .old will be deleted every time a backup is performed. This
  means only a single backup is present at any time.
  [#3354](https://github.com/memgraph/memgraph/pull/3354)
  [#3355](https://github.com/memgraph/memgraph/pull/3355)
- Optimize how the WAL file gets finalised. If the replication process isn't
  using the current WAL file, the file will be renamed instead of copied with
  the new name. Users can expect better performance, especially when importing
  data. [#3381](https://github.com/memgraph/memgraph/pull/3381)
- Vector index recovery performance has been improved by fixing an issue where
  the index specification wasn’t updated after resizing. This prevents
  unnecessary resizing during recovery and speeds up startup. No user action is
  required. [#3390](https://github.com/memgraph/memgraph/pull/3390)
- Added a script to `RelWithDebInfo` Docker images so that users can easily run
  Memgraph using `gdb`, save a core dump and output a backtrace to the terminal
  by simply overriding the entry point.
  [#3428](https://github.com/memgraph/memgraph/pull/3428)

#### 🐞 Bug fixes

- Fixed an issue where the Cartesian-to-HashJoin rewrite produced incorrect
  results for nested property lookups. Queries using nested properties will now
  always return correct results without requiring any user action.
  [#3337](https://github.com/memgraph/memgraph/pull/3337)
- Fixed incorrect list comprehension aggregation logic that caused crashes in
  certain queries. List comprehension queries now execute correctly without
  errors. No user action is required.
  [#3338](https://github.com/memgraph/memgraph/pull/3338)
- Fixing Memgraph running under k8s. It was impossible to move the file from
  `/tmp` to `/var/lib/memgraph` when Memgraph is deployed on K8s because `/tmp`
  and `/var/lib/memgraph` were on different partitions. Data directory
  (`/var/lib/memgraph`) is used through PVC, while /tmp is locally attached
  storage. The replica instance will now save incoming WAL files and snapshots
  to the `tmp` folder, located in the data directory. Previously, it was using
  `/tmp`, but then an issue arose when attempting to move files from `/tmp` to
  the data directory because the directories were located on different
  partitions. [#3359](https://github.com/memgraph/memgraph/pull/3359)
- Fixed certain list comprehensions that were previously not possible by fixing
  the input of the operator tree. Queries like `RETURN [()-->()|1];` should
  work now. [#3365](https://github.com/memgraph/memgraph/pull/3365)
- Fixed dropping of the global edge index, which wasn't properly working,
  causing a replica to crash. Users can now expect normal working when dropping
  the global edge index.
  [#3382](https://github.com/memgraph/memgraph/pull/3382)
- Fixed the case when edge indices skip deleted edges regardless of transaction
  start time and isolation level. Users who use edge indices will now see the
  correct result based on their transaction and isolation.
  [#3383](https://github.com/memgraph/memgraph/pull/3383)
- Fixed an issue of planner incorrectly discarding edge type filters when using
  global edge property indices `CREATE GLOBAL EDGE INDEX ON :(p);`. Queries
  that use a global edge index no longer ignore edge type filters.
  [#3392](https://github.com/memgraph/memgraph/pull/3392)
- Replica will now send an `InProgressRpc` message when it takes more time to
  make deltas durable. Previously, deltas might have been processed quickly,
  but due to writing to disk, the `MAIN` process thought the replica was down.
  This caused transactions to be aborted in `STRICT SYNC` mode. On the other
  hand, in `SYNC` mode, the transaction would be committed, with the exception
  being thrown, which would confuse the user about what actually happened in
  the system. [#3400](https://github.com/memgraph/memgraph/pull/3400)
- Fixed an issue where `SHOW INDEX INFO` would fail due to the presence of a
  global edge index. [#3411](https://github.com/memgraph/memgraph/pull/3411)
- Fixed an issue with operator generation in EXISTS subqueries when no
  expansion was produced, which previously could cause the database to crash.
  Such queries now execute correctly without errors.
  [#3412](https://github.com/memgraph/memgraph/pull/3412)
- Fixed decoding of the extended property value type for lists was incorrect
  because the list type was decoded multiple times: once in
  `DecodeExpectedPropertyType` and again in `SkipPropertyValue`, resulting in
  an error. [#3429](https://github.com/memgraph/memgraph/pull/3429)
- Fixed server crashing when a transaction completes by catching a
  serialization error inside a custom query module. Users should expect more
  stability because a serialization error caused by query modules shouldn't
  affect the Memgraph process.
  [#3439](https://github.com/memgraph/memgraph/pull/3439)

### MAGE v3.7.0 - November 19th, 2025

#### ✨ New features

- Added `collections.frequencies_as_map` function to count element occurrences
  in a list. [#628](https://github.com/memgraph/mage/pull/628)

#### 🛠️ Improvements

- Updated the toolchain used to build MAGE. The usages should remain the same,
  but please note that if you are building MAGE from source, you must start
  using the latest toolchain (v7).
  [#670](https://github.com/memgraph/mage/pull/670)
- Added `relwithdebinfo` builds that use `malloc` instead of `jemalloc`; also
  included `heaptrack` in all `relwithdebinfo` Docker images.
  [#681](https://github.com/memgraph/mage/pull/681)
- Added an experimental Ubuntu-24.04 DEB package to the daily build for
  installing all MAGE modules.
  [#678](https://github.com/memgraph/mage/pull/678)
- Added a script to `RelWithDebInfo` Docker images so that users can easily run
  Memgraph using `gdb`, save a core dump and output a backtrace to the terminal
  by simply overriding the entry point.
  [#693](https://github.com/memgraph/mage/pull/693)

### Memgraph v3.6.2 - November 4th, 2025

- Version bump to v3.6.2 to match the latest MAGE version.

### MAGE v3.6.2 - November 4th, 2025

#### ⚠️ Breaking changes

- Renamed `compute()` function to `node_sentence()` and simplified input
  arguments such that it now accepts a configuration `Map` to define parameters
  such as `device` and `batch_size`.
  [#686](https://github.com/memgraph/mage/pull/686)

#### ✨ New features

- Added `text()` function to embed lists of strings directly.
  [#686](https://github.com/memgraph/mage/pull/686)
- Added `model_info()` function to return a `Map` of information about the model
  being used. [#686](https://github.com/memgraph/mage/pull/686)

#### 🛠️ Improvements

- Added configuration option `return_embeddings` to `node_sentence()` so that a
  list of embeddings are returned. `node_sentence()` now returns `dimension` -
  the length of the output of the embedding model.
  [#686](https://github.com/memgraph/mage/pull/686)
- Fixed default `device` such that CPU-only containers fallback to CPU compute
  without having to specify that `device="cpu"`.
  [#686](https://github.com/memgraph/mage/pull/686)

### Lab v3.6.2 - October 31st, 2025

### Memgraph v3.6.1 - October 16th, 2025

#### ⚠️ Breaking changes

- Changed the behaviour of calls like `datetime({timezone:"Europe/Brussels"})`.
  Such calls return the current clock time. If you need the previous behaviour,
  then explicitly state one or more numeric fields as `0`, e.g,
  `datetime({year: 0, timezone:"Europe/Brussels"}`.
  [#3339](https://github.com/memgraph/memgraph/pull/3339)

#### 🐞 Bug fixes

- Fixed `convert_c.to_tree()` memory leak. Using `to_tree` will not indefinably
  increase the memory. [#3333](https://github.com/memgraph/memgraph/pull/3333)
- Fixed an issue where `SHOW PRIVILEGES` ignored the target user's database
  access rights. The `SHOW PRIVILEGES` now works correctly and will return an
  empty result if the user does not have access to the target database.
  [#3340](https://github.com/memgraph/memgraph/pull/3340)
- Fixed a cache invalidation bug that would cause the IN LIST operator to
  produce wrong results.
  [#3342](https://github.com/memgraph/memgraph/pull/3342)
- Fixed `datetime()` calls with only a `timezone` parameter (e.g.,
  `datetime({timezone:"Europe/Brussels"})`) to return the current clock time,
  instead of a time with all numeric fields set to their minimum value.
  [#3339](https://github.com/memgraph/memgraph/pull/3339)

### MAGE v3.6.1 - October 16th, 2025

#### ✨ New features

- Added `embedding` module, allowing users to compute sentence embeddings for
  vertices on CPU or Nvidia GPU.
  [#662](https://github.com/memgraph/mage/pull/662).

### Lab v3.6.1 - October 17th, 2025

### Memgraph v3.6.0 - October 8th, 2025

#### ⚠️ Breaking changes

- Removed `mg_import_csv` executable from Memgraph, `LOAD CSV` should be used
  instead. [#3216](https://github.com/memgraph/memgraph/pull/3216)
- Changed the behaviour of Memgraph when users set the
  `--storage-wal-enabled=false`, `--storage-snapshot-interval` and
  `--storage-snapshot-interval-sec` flags. If Memgraph won't start under
  replication/HA, if `--storage-wal-enabled=false`. Single instance Memgraph
  deployments should have the same behaviour.
  [#3259](https://github.com/memgraph/memgraph/pull/3259)
- Changed how MEMGRAPH_USER and MEMGRAPH_PASSWORD environment variables behave
  under coordinators. Memgraph will ignore those two environment variables,
  resulting in any hacked solution exposing this workaround and using
  bolt+routing stop to work.
  [#3278](https://github.com/memgraph/memgraph/pull/3278)
- Removed `text-search` from the list of experimental features. If you used the
  text search, please make sure to remove the
  `--experimental-enabled=text-search` because Memgraph won't be able to start.
  [#3301](https://github.com/memgraph/memgraph/pull/3301)
- `SHOW INDEX INFO` output is different. Previously, all text indexes returned
  the hardcoded string "text" as their type. Now, the output reflects the
  actual indexed entity more precisely: for node indexes `label_text`, for edge
  indexes `edge-type_text`.
  [#3178](https://github.com/memgraph/memgraph/pull/3178)

#### ✨ New features

- Support for RPC versioning has been added to enable ISSU (In-Service Software
  Upgrades). The user can now expect no-downtime upgrades when following a
  specific procedure. Before, when exchanging messages, the server could not
  read an RPC request if it was not of the latest version. Additionally,
  support has been added to update the protocol in future releases.
  [#3172](https://github.com/memgraph/memgraph/pull/3172)
- Allow update and removal of nested properties inside a node or relationship.
  Users can now perform fine-grained updates inside the map properties of nodes
  and relationships. [#3052](https://github.com/memgraph/memgraph/pull/3052)
- A new `cosine_similarity` function was added to the `vector_search` module.
  This allows users to calculate cosine similarity directly in queries, e.g.,
  `CALL vector_search.cosine_similarity([1.0, 2.0], [1.0, 2.0, 3.0])`.
  [#3240](https://github.com/memgraph/memgraph/pull/3240)
- Added `DROP ALL INDEXES` and `DROP ALL CONSTRAINTS` clauses that allow you to
  delete all indices or all constraints in your database with a single command.
  [#3247](https://github.com/memgraph/memgraph/pull/3247)
- Extended the client-side bolt+routing protocol. It will now work normally
  with multi-tenancy in HA mode. Specifying env variables MEMGRAPH_USER and
  MEMGRAPH_PASSWORD on coordinators will no longer create a user to
  authenticate. [#3278](https://github.com/memgraph/memgraph/pull/3278)
- A new coordinator setting `max_failover_replica_lag` has been added, which
  allows users to control how many transactions they are willing to risk losing
  upon a data failover. If the replica is behind main for more than
  `max_failover_replica_lag` transactions, the coordinator will forbid failover
  to such an instance. [#3219](https://github.com/memgraph/memgraph/pull/3219)
- Added `max_replica_read_lag_` coordinator setting. This new setting prevents
  routing read queries to replicas that are too far behind the current main
  instance, ensuring data consistency for read operations.
  [#3222](https://github.com/memgraph/memgraph/pull/3222)
- Allow database/tenant renaming. Users can now rename databases, instead of
  having to create a new one, recover from an old snapshot and then drop the
  old database. Simplifies admin workflow.
  [#3161](https://github.com/memgraph/memgraph/pull/3161)
- Added `FORCE` specifier to `DROP DATABASE` query. Allows users to force a
  drop even if it is currently being used. The database will be hidden
  immediately, but deletion is deferred until it is safe.
  [#3033](https://github.com/memgraph/memgraph/pull/3033)
- Updated system metrics with `WriteWriteConflicts` and `TransientErrors`
  counters [#3230](https://github.com/memgraph/memgraph/pull/3230)
- Updated list of metrics with latency of the GC skiplist cleanup, as well as
  the latency of the whole GC process
  [#3235](https://github.com/memgraph/memgraph/pull/3235)
- Added support for text indexing on edges. Users can now create and query text
  indexes on edges. The syntax for creating an edge-based text index is:
  `CREATE TEXT EDGE INDEX edge_index_name ON :EDGE_TYPE(text_property);`.
  [#3178](https://github.com/memgraph/memgraph/pull/3178)
- Added fine-grained access control for `SHOW SCHEMA INFO;` query. Users are
  now restricted from seeing the full graph schema if they don't have the
  fine-grained privileges.
  [#3304](https://github.com/memgraph/memgraph/pull/3304)

#### 🛠️ Improvements

- Added python3.12-venv inside Memgraph's build image. It makes it easier to
  create environments when using Memgraph as part of the K8s cluster.
  [#3206](https://github.com/memgraph/memgraph/pull/3206)
- Added improvement in which durability files that are transferred between
  instances in an HA environment won't require as big a memory allocation on
  the receiver's end. Before, the whole file had to be read into memory to be
  used; now we use a buffer of 64KiB to write bytes received on the network to
  the file on disk. Users with a big scale shouldn't see any memory spikes on
  replicas. This should help with Memgraph's stability when deployed in
  environments with tight memory control, e.g, K8s
  [#3227](https://github.com/memgraph/memgraph/pull/3227)
- Improved fine-grained access control error messages.
  [#3287](https://github.com/memgraph/memgraph/pull/3287)
- Enhanced routing table accuracy. Followers now forward `GetRoutingTableRpc`
  requests to the cluster leader to guarantee the correctness of the current
  routing table. This change improves routing accuracy at the expense of
  additional network
  latency.[#3222](https://github.com/memgraph/memgraph/pull/3222)
- Users can expect a slightly faster demotion to replica on the old main when
  the failover occurs. [#3128](https://github.com/memgraph/memgraph/pull/3182)
- Text search now includes a relevance score in results, a limit parameter to
  control result count, and support for aggregation queries on text edge
  indexes. These enhancements let users rank, filter, and analyze search
  results more effectively.
  [#3264](https://github.com/memgraph/memgraph/pull/3264)
- Added better log messages during user creation.
  [#3303](https://github.com/memgraph/memgraph/pull/3303)
- Updated error message about user not having the correct role-based privileges
  with more information about what the database admin should do in that case.
  [#3312](https://github.com/memgraph/memgraph/pull/3312)

#### 🐞 Bug fixes

- Fixed a bug where the current WAL file wasn't finalised on the replica when a
  new epoch was received. Users can expect fewer snapshots and WAL files being
  transferred in the case of many restarts in the cluster.
  [#3258](https://github.com/memgraph/memgraph/pull/3258)
- HA cluster was never supposed to work when flag
  `--storage-wal-enabled=false`. We now explicitly forbid starting the data
  instance when we detect such an incorrect state. The default value of
  `--storage-snapshot-interval-sec` is now 300 in all packaging scenarios. If
  both `--storage-snapshot-interval-sec` and `--storage-snapshot-interval` are
  set, the `--storage-snapshot-interval` flag will be used.
  [#3259](https://github.com/memgraph/memgraph/pull/3259)
- Fixed a concurrency bug that would cause Memgraph to get stuck rarely during
  shutdown. [#3285](https://github.com/memgraph/memgraph/pull/3285)
- Users will now be able to authenticate using SSO and bolt+routing protocol.
  Previously it was only working if directly connecting to data instances.
  [#3307](https://github.com/memgraph/memgraph/pull/3307)
- Corrupted snapshot files won't be deleted anymore. They won't be deleted
  either, but we found it incorrect to delete durability files that may contain
  valid data for a user and that are only temporarily unusable in the current
  release. [#3300](https://github.com/memgraph/memgraph/pull/3300)
- Replica will now delete the temporarily created file that is used to receive
  the WAL file replicated from the main.
  [#3330](https://github.com/memgraph/memgraph/pull/3300)

### MAGE v3.6.0 - October 8th, 2025

#### ✨ New features

- Added support for migrating data from Neo4j under the migration module.
  [#639](https://github.com/memgraph/mage/pull/639)
- Added k-nearest neighbors (KNN) algorithm implementation.
  [#676](https://github.com/memgraph/mage/pull/676)

#### 🐞 Bug fixes

- Fixed closing of database connections after migrations are completed.
  [#665](https://github.com/memgraph/mage/pull/665)

### Lab v3.6.0 - October 8th, 2025

### Memgraph v3.5.2 - September 24th, 2025

#### ✨ New features

- Added support for pattern expressions inside the where clause. Users are now
  able to do a query similar to `MATCH (a) WHERE NOT (a)-[]->(:Something)`.
  Syntax behaves identical to EXISTS patterns.
  [#3286](https://github.com/memgraph/memgraph/pull/3286)

#### 🛠️ Improvements

- Improved performance of `convert_c.to_tree`. to_tree should run some 3-4
  times faster than before.
  [#3203](https://github.com/memgraph/memgraph/pull/3203)

#### 🐞 Bug fixes

- Fixed `convert_c.to_tree`. It now merges nodes with multiple labels which
  were duplicated in the input paths. The generated tree structure is now
  correct, regardless of the number of node labels.
  [#3283](https://github.com/memgraph/memgraph/pull/3283)

### MAGE v3.5.2 - September 24th, 2025

#### 🐞 Bug fixes

- Fixed `map.merge` module to accept nullable arguments.
  [#671](https://github.com/memgraph/mage/pull/671)

### Memgraph v3.5.1 - September 11th, 2025

#### 🐞 Bug fixes

- Fixed issue where extra results were found when using `OR` label expressions
  alongside `MATCH` labels, such as `MATCH (n:A) WHERE (n:B OR n:C) RETURN n`.
  These queries now return the correct results.
  [#3237](https://github.com/memgraph/memgraph/pull/3237)
- Fixed Cypher parsing issue which would incorrectly handle any queries that
  used certain keywords, such as `Resource`, as a label. Users no longer need
  to worry about conflicts between their dataset's labels and Cypher keywords.
  [#3256](https://github.com/memgraph/memgraph/pull/3256)

### Memgraph v3.5.0 - August 27th, 2025

#### ⚠️ Breaking changes

- Transitioned from statically linking to dynamically linking `libstdc++`. This
  change may affect users running Memgraph natively, especially during
  reinstallation. If using Docker or K8s, this change shouldn't result in any
  breaking issues. [#3061](https://github.com/memgraph/memgraph/pull/3061)
- System queries (auth, replication, multi-tenant) are not targeting the
  default database "memgraph". Users must have both the correct privileges and
  access to "memgraph" to execute them. To verify user permissions on
  "memgraph", run `SHOW PRIVILEGES FOR user ON DATABASE memgraph`
  [3086](https://github.com/memgraph/memgraph/pull/3086)
- Time-to-live (TTL) durability from before v3.5.0 is ignored, and you will
  have to set up TTL again.
  [#3126](https://github.com/memgraph/memgraph/pull/3126)

#### ✨ New features

- Added `USER PROFILE` to achieve per-user resource limits and real-time
  monitoring. Admins can now control the number of active sessions, memory
  usage per user and monitor resources in real time.
  [2980](https://github.com/memgraph/memgraph/pull/2980)
- A new STRICT_SYNC replication mode is added. Its benefits are mostly
  observable in high-availability deployments where users can, when registering
  replicas as strictly sync, run a high-availability cluster without the fear
  of experiencing data loss during failovers.
  [#3026](https://github.com/memgraph/memgraph/pull/3026)
- Added Cypher support for existential/exists subqueries. Users can use the
  following Cypher construct `... EXISTS { ... subquery ... } ...`.
  [#3057](https://github.com/memgraph/memgraph/pull/3057)
  [#3120](https://github.com/memgraph/memgraph/pull/3120)
- Added support for multiple roles per user, enabling database-specific
  permissions and enhanced SSO integration. Users can now have different roles
  for different databases, and SSO providers can assign multiple roles that
  will all be recognized by the system. Existing deployments will
  automatically migrate. New multi-role features are introduced, e.g., `SHOW
  CURRENT ROLE` and enhanced `SET ROLE/CLEAR ROLE` commands.
  [3086](https://github.com/memgraph/memgraph/pull/3086)
- Query modules now support timezone-aware datetime objects. This means your
  custom procedures and functions can now accept and return datetime values
  that include timezone information in either named IANA or numeric offset
  formats. [#3121](https://github.com/memgraph/memgraph/pull/3121)
  [#3200](https://github.com/memgraph/memgraph/pull/3200)
- Added methods `SHOW INDEXES`, `SHOW CONSTRAINTS`, `SHOW VECTOR INDEXES`,
  `SHOW ACTIVE USERS`, `SHOW TRIGGER INFO` for better compatibility and
  robustness when operating Memgraph.
  [#3149](https://github.com/memgraph/memgraph/pull/3149)
- Added the ability to create a text index on a subset of properties using the
  following syntax: `CREATE TEXT INDEX index_name ON :Label(prop1, prop2,
  prop3);` [#3155](https://github.com/memgraph/memgraph/pull/3155)
- `CREATE SNAPSHOT` returns the path to the newly created file. The query will
  not fail due to no new changes present (snapshot will be forced when the user
  requests it). Users can capture the return value and create a more complex
  snapshot handling logic on top of Memgraph.
  [3166](https://github.com/memgraph/memgraph/pull/3166)
- A new query, `SHOW REPLICATION LAG;` has been added and can be executed on
  the cluster's leader. Users can now see how many transactions were committed
  on each instance and use this information to decide whether to perform a
  failover or not. [#3167](https://github.com/memgraph/memgraph/pull/3167)
- Added [KShortest deep path
  traversal](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal)
  using the `((n1)-[*KShortest]->(n2))` syntax. The implementation offers
  shortest-path search between two nodes with path length filtering, hop
  limits, and result limits. Users can now query multiple shortest paths,
  returned in order of length, and restrict them by length, hops, or number of
  results. Use `WITH n1, n2` to define the endpoints, and apply filters or
  result limits directly in the KShortest expression as needed.
  [3170](https://github.com/memgraph/memgraph/pull/3170)
- Added `convert_c.to_tree` procedure. The procedure transforms a collection of
  `PATH` objects into a structured, nested document - a "tree" of maps -
  representing one or more root entities and their connected relationships in
  hierarchical form. `to_tree` is equivalent to Neo4j's `apoc.convert.toTree`.
  [#3089](https://github.com/memgraph/memgraph/pull/3089)
  [#3148](https://github.com/memgraph/memgraph/pull/3148)
  [#3158](https://github.com/memgraph/memgraph/pull/3158)
  [#3165](https://github.com/memgraph/memgraph/pull/3165)
  [#3185](https://github.com/memgraph/memgraph/pull/3185)

#### 🛠️ Improvements

- Several durability-related issues under text search have been fixed, and
  replication support has been added. Text search queries now correctly handle
  boolean logic and uppercase letters. Users can now safely recover data from
  databases. Replication can be used if the text search flag is enabled on both
  the main and replica. [#3095](https://github.com/memgraph/memgraph/pull/3095)
- Added missing query module mappings such as `algo`, `convert`, `coll`,
  `export`, `math`, `node`, `path`, `refactor`, `text` and `util` under
  `/etc/memgraph/apoc_compatibility_mappings.json`. Queries using those APOC
  functionalities don't have to be refactored. That's especially useful when
  GraphQL middleware/transpiler is used in front of Memgraph.
  [#3109](https://github.com/memgraph/memgraph/pull/3109)
  [#3128](https://github.com/memgraph/memgraph/pull/3128)
  [#3143](https://github.com/memgraph/memgraph/pull/3143)
  [#3152](https://github.com/memgraph/memgraph/pull/3152)
  [#3159](https://github.com/memgraph/memgraph/pull/3159)
  [#3162](https://github.com/memgraph/memgraph/pull/3162)
- Added information about database UUID under `SHOW STORAGE INFO`.
  [#3111](https://github.com/memgraph/memgraph/pull/3111)
- Added more logs when authenticating via SSO for an easier debugging
  experience. [#3113](https://github.com/memgraph/memgraph/pull/3113)
- `SHOW DATABASE` does not require any special privileges. Users can query
  their current database regardless of whether they have `MULTI_DATABASE`
  privileges. [3123](https://github.com/memgraph/memgraph/pull/3123)
- Removed the first line showing the next scheduler snapshot from `SHOW
  SNAPSHOT;` instead added a special query `SHOW NEXT SNAPSHOT`. Users can now
  explicitly query information regarding the next scheduled snapshot.
  [3123](https://github.com/memgraph/memgraph/pull/3123)
- General improvements to `oidc.py` auth module.
  [3123](https://github.com/memgraph/memgraph/pull/3123)
- Auto-index creation (when enabled) now runs asynchronously in dedicated
  background transactions and takes advantage of the new concurrent index
  creation. The commit stage is no longer blocked by auto-index creation. This
  change improves database performance and reliability.
  [#3124](https://github.com/memgraph/memgraph/pull/3124)
- Added time-to-live (TTL) improvements. TTL now supports replication.
  Internally, we also now use concurrent index creation, meaning TTL query
  commands block for less time.
  [#3126](https://github.com/memgraph/memgraph/pull/3126)
- [All shortest paths
  traversal](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#all-shortest-paths)
  should be roughly 5% faster.
  [#3127](https://github.com/memgraph/memgraph/pull/3127)
- Improved recovery under HA while loading the WAL file after the instance's
  restart. The recovery will check whether the given WAL actually contains some
  changes before saving the WAL's epoch to the history. Users could benefit
  from it since they will see less frequent brute-force recovery of the replica
  and more incremental recovery with only the data the replica needs.
  [#3141](https://github.com/memgraph/memgraph/pull/3141)
- Pattern comprehension support was added under the `WHERE` clause. It's
  possible to run queries like `MATCH (a) WHERE single(x in [(a)-[:R]->(b)
  WHERE b IS NOT null | 1] WHERE true) RETURN a;`. It's also possible to nest
  pattern comprehension under the `WHERE` clause.
  [#3146](https://github.com/memgraph/memgraph/pull/3146)
  [#3169](https://github.com/memgraph/memgraph/pull/3169)
- Reduced Docker image size by ~100MB, meaning more hard drive space for cat
  memes. [#3151](https://github.com/memgraph/memgraph/pull/3151)
- Improved `date.convert_format` function when `null` value is provided by not
  making an exception and returning null instead.
  [#3157](https://github.com/memgraph/memgraph/pull/3157)
- The `mgcxx` version has been updated to `0.0.9`, bringing drop index
  enhancements and text search performance improvements. For users, this means
  faster index operations, and `SHOW INDEX INFO` now returns the number of
  entries in a text index instead of null.
  [#3171](https://github.com/memgraph/memgraph/pull/3171)
- Make OIDC SSO work without the ID token if it's not needed. Users
  authenticating with OIDC SSO via a Neo4j driver don't need to pass an ID
  token if they don't use it for the username field.
  [#3175](https://github.com/memgraph/memgraph/pull/3175)

#### 🐞 Bug fixes

- The query memory tracker has been moved to custom malloc/free
  implementations, allowing more precise tracking. To ensure modules use the
  correct implementation, libstd is dynamically linked. Users must ensure they
  have the correct libstd version (in the case Memgraph is natively installed).
  [3061](https://github.com/memgraph/memgraph/pull/3061)
- [All shortest paths
  traversal](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#all-shortest-paths)
  was missing some paths. It correctly returns all results now.
  [#3119](https://github.com/memgraph/memgraph/pull/3119)
- Fixed a bug when the query module is called inside another query module.
  Query modules can now be safely nested without causing execution errors.
  Users can now structure complex logic using nested query modules like `CALL
  do.when(...)` without workarounds.
  [#3120](https://github.com/memgraph/memgraph/pull/3129)
- Fixed an issue where SSO users were missing from the `SHOW ACTIVE USERS INFO`
  results. Admins can now see a complete list of logged-in users.
  [3144](https://github.com/memgraph/memgraph/pull/3144)
- Fixed a bug in the `UNWIND` clause shadowing symbols in particular
  subqueries. [#3154](https://github.com/memgraph/memgraph/pull/3154)
- Fixed a concurrency issue in text index updates. From the user's perspective,
  text index operations will now run more reliably under concurrent workloads.
  [#3177](https://github.com/memgraph/memgraph/pull/3177)
- Changing your own password does not require any particular privilege. Users
  can now change their own passwords even if they do not have the AUTH
  privilege. [3192](https://github.com/memgraph/memgraph/pull/3192)
- Resolved segmentation fault in `AllShortestPathCursor` due to incorrect
  deletion of the last element from the vector (internal implementation
  detail). Users should no longer encounter segfaults when executing all
  shortest path queries.
  [#3202](https://github.com/memgraph/memgraph/pull/3202)
- `MonotonicBufferResource` allocate method now checks for possible overflow.
  Users can expect more accurate memory measurements.
  [#3204](https://github.com/memgraph/memgraph/pull/3204)

### MAGE v3.5.0 - August 27th, 2025

#### ✨ New features

- Added `collections_module.flatten` function for flattening nested
  collections. Example usage: `RETURN collections_module.flatten([[1, 2], [3,
  4], 5]);`. [#624](https://github.com/memgraph/mage/pull/624)
- Added `text.indexOf` function. An example of usage is `RETURN
  text.indexOf("string", "what", from, to);`.
  [#645](https://github.com/memgraph/mage/pull/645)
  [#648](https://github.com/memgraph/mage/pull/648)
- Added `math.round` function for rounding numbers with various modes.
  [#652](https://github.com/memgraph/mage/pull/652)

#### 🛠️ Improvements

- Added normalization of UTF8 characters in the `text.distance` module, making
  the module work for inputs that are UTF8 strings.
  [#640](https://github.com/memgraph/mage/pull/640)
- Added optimization for `text.regreplace` by caching the regex pattern,
  resulting in faster execution.
  [#641](https://github.com/memgraph/mage/pull/641)
- Improved the case when null is passed to the `json_util.from_json_list`
  function by just returning null in that case.
  [#655](https://github.com/memgraph/mage/pull/655)

#### 🐞 Bug fixes

- Fixed handling the specific MySQL types (Blob, Geometry, etc.) so the
  migration does not fail with the `Unsupported PyObject conversion`.
  [#656](https://github.com/memgraph/mage/pull/656)

### Lab v3.5.0 - August 27th, 2025

### Memgraph v3.4.0 - July 10th, 2025

#### ⚠️ Breaking changes

- Introduced a small breaking change under `SHOW INDEX INFO` output.
  Previously, all vector indexes returned the hardcoded string `"vector"` as
  their type. Now, the output reflects the actual indexed entity more
  precisely: for node indexes the type is `label+property_vector`, for edge
  indexes: `edge-type+property_vector`.
  [#3059](https://github.com/memgraph/memgraph/pull/3059)

#### ✨ New features

- Quantization options have been added to the vector index. Users can now
  reduce memory consumption by enabling quantization in the config map, using
  the `scalar_kind` field to define the scalar type for vectors, at the cost of
  some precision. [#3037](https://github.com/memgraph/memgraph/pull/3037)
- Added function `getHopsCounter()` to return the current number of hops
  traversed in a given query.
  [#3048](https://github.com/memgraph/memgraph/pull/3048)
- Added monitoring of replica recovery process. Users will now be able to
  monitor how often an instance gets into recovery replication state and how
  often instance gets successfully recovered once in that state.
  [#3002](https://github.com/memgraph/memgraph/pull/3002)
- Added support for vector indexing on edges (edge_type & property
  combination). Users can now create and query vector indexes on edge
  embeddings. The syntax for creating an edge-based vector index is: `CREATE
  VECTOR EDGE INDEX vector_index_name ON :EDGE_TYPE(embedding) WITH CONFIG
  config_map;`. [#3059](https://github.com/memgraph/memgraph/pull/3059)

#### 🛠️ Improvements

- Added support for `@neo4j/graphql` middleware v7.2.0. Users can now leverage
  better Neo4j / GraphQL compatibility and use Memgraph with the latest GraphQL
  ecosystem. [#3027](https://github.com/memgraph/memgraph/pull/3027)
  [#3083](https://github.com/memgraph/memgraph/pull/3083)
- Added non-blocking index creation. Until this point, index creation was a
  "stop the world" operation on a given database. With this change, at the time
  of `CREATE INDEX` queries, memgraph will exhaust all in-flight write
  operations and pause/queue all new write operations. When all in-flight
  writes are done, Memgraph will make the index usable and allow writing. Read
  queries are possible at any time. From the user perspective, `CREATE INDEX`
  operations should feel completely non-blocking. Only in extreme cases under
  heavy-write load might there be some pausing of write queries, but none of
  the writes should be lost.
  [#2923](https://github.com/memgraph/memgraph/pull/2923)
  [#3064](https://github.com/memgraph/memgraph/pull/3064)
  [#3073](https://github.com/memgraph/memgraph/pull/3073)
  [#3091](https://github.com/memgraph/memgraph/pull/3091)

#### 🐞 Bug fixes

- Queries using regex clauses on label+property indices no longer return an
  empty results set. Users can now use label+property indices to improve
  performance of regex queries (although note that, unlike exact text match
  queries, regex queries require an additional filtering step).
  [#3025](https://github.com/memgraph/memgraph/pull/3025)
- Queries now support directly accessing properties of maps nested inside
  lists, such as `RETURN [{name: 'Alice'}, {name: 'Bob'}][1].name;`. This
  allows users to more easily work with Memgraph if their schema contains
  nested data. [#3058](https://github.com/memgraph/memgraph/pull/3058)
- Fixed HA upgrade. Upgrade of coordinators from 3.2.1 to 3.3 was failing due
  to a bug in the serialisation code. With this bug fix, users can expect
  smooth upgrade of coordinators.
  [#3078](https://github.com/memgraph/memgraph/pull/3078)
- Executing `RECOVER SNAPSHOT` on `MAIN` will propagate the snapshot to
  replicas as well. This simplifies the recovery of the whole cluster. Use
  `RECOVER SNAPSHOT` to easily overwrite the data stored in a database.
  [3031](https://github.com/memgraph/memgraph/pull/3031)

### MAGE v3.4.0 - July 10th, 2025

#### ✨ New features

- Added `text.replace`, `text.regreplace` and `text.distance` functions.
  [#621](https://github.com/memgraph/mage/pull/621)
- Added `refactor.mergeNodes` procedure.
  [#623](https://github.com/memgraph/mage/pull/623)
- Added `date.convert_format()` function, which converts temporal strings in
  one format to another temporal string format.
  [#631](https://github.com/memgraph/mage/pull/631)

### Lab v3.4.0 - July 16th, 2025

### Memgraph v3.3.0 - June 4th, 2025

#### ✨ New features

- New bolt work scheduler based on a priority queue with a sidecar channel for
  high priority queries. Users should never see a blocked database due to heavy
  load. Users can now run specific high priority queries even under heavy loads
  (example: SHOW TRANSACTIONS).
  [#2567](https://github.com/memgraph/memgraph/pull/2567)
- Added support for creating indices on nested properties within maps. Users
  can now create efficient indices on deeply nested data structures using dot
  notation, such as `CREATE INDEX ON :User(preferences.notifications.email)`,
  eliminating the need to flatten complex nested properties for query
  performance.** [#2934](https://github.com/memgraph/memgraph/pull/2934)
- Establishing the connection between sockets now has a timeout of 5s. Users
  can expect lower and more predictable p99 latencies for the time needed to
  establish a socket connection. The timeout for StateCheckRpc and
  FrequentHeartbeatRpc is reduced to 5s. This will help heartbeat mechanisms to
  figure out faster when the instance is down or when some steps need to be
  taken to bring replica to the same level on which the main is.
  [#2949](https://github.com/memgraph/memgraph/pull/2949)
- Memgraph now exposes a new set of metrics on HA cluster: how long the data
  failover takes and how long starting and finalising transaction takes on
  replica instances. [#2965](https://github.com/memgraph/memgraph/pull/2965)

#### 🛠️ Improvements

- Updated `PropertyValue` structure to use PropertyId keys for improved
  performance, reduced memory usage, and simplified handling. Introduced
  `ExternalPropertyValue` type with string-keyed map for human-readable
  property names in parsing/serialization contexts.
  [#2901](https://github.com/memgraph/memgraph/pull/2901)
- Added daily build of `RelWithDebInfo` Docker image for `arm64`. Allows ARM
  users to debug bleeding edge Memgraph builds.
  [#2996](https://github.com/memgraph/memgraph/pull/2996)

#### 🐞 Bug fixes

- Fixed memory leaks in several C++ API functions, including those related to
  vector search, text search, index management, and constraints.**
  [#2934](https://github.com/memgraph/memgraph/pull/2934)
- TTL, streams and CSV import could fail when using auth modules. When using
  auth modules, we require a user/role. In these specific cases we want an
  empty user, which fails on creation. Users can now use stream, TTL and init
  from cypher file with auth modules.
  [2967](https://github.com/memgraph/memgraph/pull/2967)
- `utils::Scheduler` wouldn't use strict increments when tasks took longer than
  the period. This would cause the start time to drift. Users could have seen
  this when running TTL on v3.2. Now the start time will be respected. In cases
  where tasks take longer, the next execution will be immediate, but subsequent
  will follow the schedule.
  [2970](https://github.com/memgraph/memgraph/pull/2970)
- The asynchronous replication is improved for situations in which the recovery
  task would've finished while another transaction is getting replicated.
  Previously, such transaction would never get replicated on async replica
  which wasn't the intended
  behaviour.[#2973](https://github.com/memgraph/memgraph/pull/2973)
- When an instance becomes main, it marks a range of transactions with IDs
  smaller than global timestamp used in replication as finished. This should
  ensure consistent memory behaviour both on main and replicas. In the same
  way, transactions will be also marked finished when SnapshotRpc is received
  or when the `RECOVER SNAPSHOT` query is used on the standalone Memgraph.
  [#2976](https://github.com/memgraph/memgraph/pull/2976)
- Tasks for checking replica state and recovery task can now be skipped if it
  is impossible to get unique access to RPC lock for ASYNC replicas. This is to
  ensure that the commit task executes as soon as possible so that the deadlock
  between commit and recover tasks cannot happen anymore. There are no UX-wise
  changes but users can expect more reliable ASYNC replication.
  [#2984](https://github.com/memgraph/memgraph/pull/2984)
- Main instance won't block while trying to acquire RPC stream for replicating
  txn to ASYNC replica. It will instead consider replica as maybe behind. User
  can expect deadlock-free behavior when using ASYNC replica. Commits will
  happen faster while the replication lag will be a bit bigger.
  [#2989](https://github.com/memgraph/memgraph/pull/2989)
- Made parallel snapshot more flexible while combining temporary file into a
  single snapshot. Users are now less restricted on the batch size they can
  use. [2995](https://github.com/memgraph/memgraph/pull/2995)
- Fixed queries using `WHERE ... IN` clauses no longer return an empty results
  set when using label+property indices.
  [#3014](https://github.com/memgraph/memgraph/pull/3014)
- Fixed memory leak that occurred when queries updated deeply nested
  properties. Users can now safely create queries that update nested map
  properties without risk of eventual out-of-memory errors.
  [#3016](https://github.com/memgraph/memgraph/pull/3016)

### MAGE v3.3.0 - June 4th, 2025

#### 🛠️ Improvements

- Upgrade torch version to 2.6 to address [this
  vulnerability](https://github.com/pytorch/pytorch/security/advisories/GHSA-53q9-r3pm-6pq6).
  [#614](https://github.com/memgraph/mage/pull/614)

### Lab v3.3.0 - June 4th, 2025

### Memgraph v3.2.1 - May 9th, 2025

#### 🛠️ Improvements

- Modified the Docker image such that the `memgraph` user has the UID `101` to
  match that used by `MAGE`. If the user has any volumes mounted on older
  versions or memgraph (< 3.2.1), then they should `chown 101:101
  /path/to/volume` to allow the `memgraph` user to read and write to them.
  [#2947](https://github.com/memgraph/memgraph/pull/2947)
- New `relwithdebinfo` Docker build for `arm64` to allow users to debug
  Memgraph on ARM. Push `relwithdebinfo` DEB packages and Docker images to
  download bucket during release, so they are available
  [here](https://memgraph.com/download)
  [#2952](https://github.com/memgraph/memgraph/pull/2952)
- Improved the performance when writing to disk log stores' durable data using
  batched writing. Users can expect the same behavior as before, with a
  slightly better performance when using high availability.
  [#2946](https://github.com/memgraph/memgraph/pull/2946)

#### 🐞 Bug fixes

- Fixed `PageSlabMemoryResource` large allocation returning wrong address.
  Creating deltas with large properties would correctly allocate memory, but
  cause an out-of-bounds write. Users shouldn't see segfaults on commits or GC
  runs anymore. [#2950](https://github.com/memgraph/memgraph/pull/2950)
- Fixed CallProcedureCursor reset method. Query modules would sometimes cause a
  crash due to not being reset properly.
  [#2953](https://github.com/memgraph/memgraph/pull/2953)
- Fixed impersonate user misalignment between interpreter and session. This
  broke the permissions when using the impersonated user. User can now actually
  use the feature and see the correct behavior.
  [#2938](https://github.com/memgraph/memgraph/pull/2938)
- Fixed query type deduction giving the wrong answer in cases where subqueries
  execute read procedures. Without the fix users could not execute certain
  queries. Query type misalignment shouldn't occur anymore.
  [#2944](https://github.com/memgraph/memgraph/pull/2944)
- Improved usages of read-modify-write pattern in the Raft code by using compare
  exchange operations instead of blindly loading, modifying and then storing
  again. Users should expect the same behaviour from the perspective of
  correctness but with a smaller chance of something getting inconsistent.
  [#2937](https://github.com/memgraph/memgraph/pull/2937)

### Memgraph v3.2.0 - Apr 23rd, 2025

#### ⚠️ Breaking changes

- Procedure results are now populated with null values by default. Procedures
  no longer throw if all result fields don't get inserted; instead, they return
  null in those fields. [#2790](https://github.com/memgraph/memgraph/pull/2790)

#### ✨ New features

- Added support for composite indices. Composite indices are indices on labels
  and two or more properties. Composite indices significantly improve database
  performance on queries that filter against multiple properties. To take
  advantage of this improvement, users should create composite indices for
  frequently queried combinations of properties.
  [#2760](https://github.com/memgraph/memgraph/pull/2760),
  [#2887](https://github.com/memgraph/memgraph/pull/2887)
- Added metrics tracking related to replication and high availability. This
  includes metrics like the number of RPC messages sent, the duration of the
  recovery etc. Users can use our
  [prometheus-exporter](https://github.com/memgraph/prometheus-exporter) to get
  immediately cluster info status.
  [#2772](https://github.com/memgraph/memgraph/pull/2772)
- Added `mgp_result_reserve` function to the C-api for reserving memory for
  procedure results. [#2790](https://github.com/memgraph/memgraph/pull/2790)
- Added edge TTL. The system can now automatically delete stale edges using the
  specified property `ttl` inside the edges.
  [#2730](https://github.com/memgraph/memgraph/pull/2730)
- Added global index on edge properties. Since edges have only one relationship
  type, the global property index will have a data structure that is tracked
  over properties with the syntax `CREATE GLOBAL EDGE INDEX ON :(prop)`.
  [#2730](https://github.com/memgraph/memgraph/pull/2730)
- Expanded storage access types to UNIQUE, READ, WRITE and READ_ONLY. A better
  differentiation of query types allows more queries to run in parallel. Users
  can now run read queries while creating snapshots, even in ANALYTICAL mode.
  [2798](https://github.com/memgraph/memgraph/pull/2798)
- Added support for OR label expressions, enabling queries like `MATCH
  (n:Label1|Label2)` to retrieve nodes with any of the specified labels. If
  indexes exist on the labels, the planner rewrites the query as a UNION of
  index scans for efficient execution. Additionally, simple WHERE clauses with
  OR label checks are optimized similarly, reducing reliance on full scans.
  [#2783](https://github.com/memgraph/memgraph/pull/2783)

#### 🛠️ Improvements

- There is now an improved logic for selecting the new MAIN instance during the
  failover in the MT environment. An instance that has the newest databases is
  considered the newest. Users will now have a smaller chance of experiencing
  the loss of committed data.
  [#2812](https://github.com/memgraph/memgraph/pull/2812)
- Metrics related to HA coordinators are now immediately reset after the client
  pulls them. Users can now see the aggregated value from all coordinators when
  using [Memgraph's Prometheus
  exporter](https://github.com/memgraph/prometheus-exporter/pull/19). Also, the
  data won't be lost after the coordinator's restart.
  [#2802](https://github.com/memgraph/memgraph/pull/2802)
- The memory footprint of TypedValue (internal data structure used mostly
  during query execution) got smaller. The optimization results in lower memory
  usage in query modules that return many results (e.g., modules that return
  the whole graph, like PageRank) by a few percent.
  [#2806](https://github.com/memgraph/memgraph/pull/2806)
- Reduced memory footprint of `mgp_result_record` used for returning results in
  all procedures APIs. The optimization reduces memory usage by up to 50% in
  query modules, which return many results, such as the whole graph (e.g.,
  PageRank). [#2790](https://github.com/memgraph/memgraph/pull/2790)
- `utils::Scheduler` now skips backed-up tasks. Background tasks that took
  longer than the period will not be called multiple times after the execution.
  It lowers the overall workload and frees up resources.
  [2853](https://github.com/memgraph/memgraph/pull/2853)
- Schema info edge updates made faster. Previously, edge deletions would cause
  a slow commit. Users can now use the run-time schema info more freely without
  impacting the performance.
  [2844](https://github.com/memgraph/memgraph/pull/2844)
- Snapshot creation has been parallelized. Edges and vertices are batched (as
  defined via flag). These batches are processed in parallel on a number of
  threads. Users can now create snapshots in less time, freeing up resources
  for other tasks. [2818](https://github.com/memgraph/memgraph/pull/2818)

#### 🐞 Bug fixes

- Parsing a query in which a pattern comprehension is nested (such as `RETURN
  [[(x)-->(y) | x]];`) no longer causes a crash.
  [#2716](https://github.com/memgraph/memgraph/pull/2716)
- Replica won't crash anymore when performing recovery and when disk access to
  the durability folder fails. Instead, it will return an empty response,
  signaling the main that the recovery isn't completed. Users should expect the
  same behavior as before.
  [#2780](https://github.com/memgraph/memgraph/pull/2780)
- The scheduler (utility module) will release its lock before starting to
  execute its function. This will enable deadlock-free concurrent execution
  between the periodic snapshot thread and the thread that performs demotion
  under replication. [#2882](https://github.com/memgraph/memgraph/pull/2882)
- During the promotion of the MAIN instance, the promotion will be aborted if
  registering one of the replicas fails. In that case, the coordinator will
  retry the promotion after a few seconds. There is now a smaller chance of
  users experiencing failed failover. Previously, the RPC communication would
  fail, and MAIN would partially register some of the replicas.
  [#2869](https://github.com/memgraph/memgraph/pull/2869)
- The newly elected coordinator leader will perform a failover correctly when
  the leader inherits the cluster state without the main instance.
  [#2857](https://github.com/memgraph/memgraph/pull/2857)
- When manually promoting an instance, it is now unnecessary that all replicas
  are alive. Users can now use the `SET INSTANCE instance TO MAIN` query in
  more situations. [#2857](https://github.com/memgraph/memgraph/pull/2857)
- If a snapshot gets created while MAIN receives a request to demote itself to
  REPLICA, the snapshot creation will get aborted. Additionally, the thread
  creating snapshots will be paused on demotion. Similarly, when the REPLICA is
  promoted to the main, the snapshot thread will be resumed.
  [#2846](https://github.com/memgraph/memgraph/pull/2846)
- The `DUMP DATABASE` query generated the wrong query for ANY triggers. Dumping
  the database now generates the correct trigger query.
  [2855](https://github.com/memgraph/memgraph/pull/2855)
- Fixed missing edge property index and unique constraint garbage collection.
  No slow memory creep up while using constraints or edge index.
  [2850](https://github.com/memgraph/memgraph/pull/2850)
- Metadata from WAL files was unnecessarily used during recovery, even for old
  WAL files which shouldn't have been used. This could've caused a possible
  data loss because newer WAL files wouldn't be used in some situations while
  recovering. [#2892](https://github.com/memgraph/memgraph/pull/2892)
- `DUMP DATABASE` has more stability guarantees. The output order has changed
  for better memory usage and performance, and a sort is applied to the type
  constraints. The durability write bug in v2.19.0, v2.20.0, v2.20.1 for the
  3D Cartesian data type was fixed in v2.21.0, but in a way that reading
  durability from those versions would no longer be possible if a "corrupt"
  3D Cartesian was written into those durability files. Those durability
  files can now be read, and the engine handles the "corrupt" values.
  [#2888](https://github.com/memgraph/memgraph/pull/2888)
- Fixed unreliable authentication deserialization. Users can now reliably
  upgrade auth data between different versions, as well as between the
  community and enterprise.
  [2915](https://github.com/memgraph/memgraph/pull/2915)

### MAGE v3.2.0 - Apr 23rd, 2025

#### ⚠️ Breaking changes

- Significantly improved image build time and size -> there is only "with-ML"
  image now -> please use that one instead of previous no-ML images.
  [#590](https://github.com/memgraph/mage/pull/590)
- Shorten Docker image tags e.g. from `3.2-memgraph-3.2` -> `3.2` or
  `3.2-memgraph-3.2-relwithdebinfo` -> `3.2-relwithdebinfo` for all future
  releases. Users updating to images for Memgraph/MAGE >= 3.2 should use the
  shorter tag format. [#600](https://github.com/memgraph/mage/pull/600)

#### ✨ New features

- Added migration module via the Arrow Flight RPC protocol. Users can now
  stream data into Memgraph using a single Cypher query by executing the `CALL
  migrate.arrow_flight()` procedure.
  [#558](https://github.com/memgraph/mage/pull/558)
- Added migration module using `DuckDB`. Users can now connect to their data
  sources with DuckDB using a single Cypher query by executing the `CALL
  migrate.duckdb()` procedure.
  [#560](https://github.com/memgraph/mage/pull/560)
- Added the migration module from another Memgraph instance. Users can now
  seamlessly transfer data from one Memgraph instance to another using a single
  Cypher query by executing the `CALL migrate.memgraph()` procedure.
  [#561](https://github.com/memgraph/mage/pull/561)
- Added migration module to migrate from `ServiceNow` data source. Users can
  now migrate their data using ServiceNow REST API by executing a single Cypher
  query with `CALL migrate.servicenow()` procedure
  [#572](https://github.com/memgraph/mage/pull/572)

#### 🛠️ Improvements

- The performance of the betweenness centrality and online betweenness
  centrality algorithms has been improved, with speedups exceeding 100% in some
  cases. By default, the algorithm now uses half of the available hardware
  threads for parallelization.
  [#588](https://github.com/memgraph/mage/pull/588)
- Updated torch versions to 2.4.0 to fix
  [CVE-2024-7804](https://github.com/advisories/GHSA-4vmg-rw8f-92f9); use
  cached Python packages for faster CI builds.
  [#587](https://github.com/memgraph/mage/pull/587)

#### 🐞 Bug fixes

- Moved `$HOME` for `memgraph` user to fix missing Python modules when users
  mount `/var/lib/memgraph` as an external volume.
  [#594](https://github.com/memgraph/mage/pull/594)

### Lab v3.2.0 - Apr 23rd, 2025

### Memgraph v3.1.1 - Mar 28th, 2025

#### ✨ New features

- Added list comprehension Cypher feature. It's now possible to execute queries
  like `MATCH (x) RETURN [(x)-->(y) | y.prop];`.
  [#2656](https://github.com/memgraph/memgraph/pull/2656)
- Added `--storage-access-timeout-sec` flag. Previously the access timeout was
  hardcoded to 1sec. Now users can define an appropriate timeout for their
  workflow at startup. [#2831](https://github.com/memgraph/memgraph/pull/2831)

#### 🛠️ Improvements

- Added OIDC SSO quality of life change. Previously, the first role from the
  roles field would always get picked even if it wasn't in the mappings. Allow
  multiple roles to be present in the roles field, but only if all those are in
  role mappings and mapped to the same Memgraph role.
  [#2799](https://github.com/memgraph/memgraph/pull/2799)
- Allow redirecting to nonlocalhost and HTTPS in SAML SSO. `CALLBACK_URL` can
  now use HTTPS and isn't hardcoded to the localhost.
  [#2799](https://github.com/memgraph/memgraph/pull/2799)

#### 🐞 Bug fixes

- Timeout for `SwapMainUUIDRpc` message is now properly configured with the
  default value set to 10s. Users shouldn't expect any difference in the
  replication behavior. [#2826](https://github.com/memgraph/memgraph/pull/2826)
- Install correct SAML dependencies under MAGE. SAML SSO should now work
  correctly MAGE. [#2829](https://github.com/memgraph/memgraph/pull/2829)
- `EnableWritingOnMainRpc` will now be sent when the failover occurs, even if
  the network identity of some of the replicas isn't stable. Users will now be
  able to test HA in a docker-compose environment.
  [#2817](https://github.com/memgraph/memgraph/pull/2817)
- Fixed deadlock caused by write queries and periodic snapshots.
  [2819](https://github.com/memgraph/memgraph/pull/2819)
- Broken migration from old durable auth data. Migration is now fixed, and
  pre-v3.1 auth data can be correctly read. Users can upgrade to v3.1 without
  encountering authentication problems.
  [2820](https://github.com/memgraph/memgraph/pull/2820)
- RPC timeouts for messages requesting promotion, demotion, enabling writing on
  main and databases' history will now work correctly. Users shouldn't observe
  long-running RPCs in their HA environment anymore.
  [#2813](https://github.com/memgraph/memgraph/pull/2813)
- Memgraph DB doesn't know how it will get deployed on K8s. It is possible to
  abstract multiple coordinators under the same load balancer. In that case,
  the Bolt server is the same for all coordinators. Previously, the code didn't
  allow adding multiple coordinators with the same Bolt server. Now, the users
  will be able to do that.
  [#2801](https://github.com/memgraph/memgraph/pull/2801)
- Fix the segfault issue caused by a double free. It only happens when Python
  procedures are called in IN_MEMORY_ANALYTICAL mode.
  [#2804](https://github.com/memgraph/memgraph/pull/2804)
- The serialization reader would throw if data from the buffer weren't fully
  ready. That caused network connections between the main and replica to be
  constantly dying. Now, when the replica doesn't read all data from the
  buffer, the exception thrown in the SLK part of the code is handled properly,
  avoiding the connection drop. Additionally, the PR tries to reduce the
  massive amount of logs produced in HA code when trace log level is used.
  [#2770](https://github.com/memgraph/memgraph/pull/2770)
- `SnapshotRpc`, `WalFilesRpc`, and `CurrentWalRpc` will now wait for a main
  lock but for a maximum time of 30s. If the lock cannot be acquired within the
  30s on the replica, the main try to replicate again after a few seconds. Such
  timeout is added to prevent a possible deadlock when multiple fast failovers
  happen during transaction execution while the recovery process is in place.
  Users should observe the same behavior as before.
  [#2816](https://github.com/memgraph/memgraph/pull/2816)

### MAGE v3.1.1 - Mar 28th, 2025

#### ✨ New features

- Added migration module from Neo4j. Users can now seamlessly insert data by
  using one single Cypher query to migrate their data from Neo4j. Query is
  performed with the `CALL migrate.neo4j()` procedure
  [#557](https://github.com/memgraph/mage/pull/557)

#### 🛠️ Improvements

- Added `gdb` to and full MAGE repo source to the RelWithDebInfo Docker image.
  This will allow for debugging instances without requiring a connection to the
  internet [#579](https://github.com/memgraph/mage/pull/579)
- Reduced image size by cleaning up after installing packages.
  [#568](https://github.com/memgraph/mage/pull/568),
  [#564](https://github.com/memgraph/mage/pull/564)

#### 🐞 Bug fixes

- Set default shell for `memgraph` user to `/bin/bash` for consistency with all
  previous Docker images. This should not change anything from the user's
  perspective [#577](https://github.com/memgraph/mage/pull/577)
- The `export_util` query module didn't specify the export encoding. If your
  data had special characters not supported by ASCII, they would be saved as
  garbage in the exported file, or the export would fail.  We have fixed this
  by predefining the UTF-8 encoding in the export utility to support special
  characters [#554](https://github.com/memgraph/mage/pull/554)

### Memgraph v3.1.0 - Mar 12th, 2025

#### ⚠️ Breaking changes

- Improved error messages might result in a `breaking` outcome under your
  applications if the application assumes the exact content of error messages.
  [#2379](https://github.com/memgraph/memgraph/pull/2379)
- `include/mg_exceptions.hpp` had ABI break which requires C++ modules to be
  rebuilt. [#2702](https://github.com/memgraph/memgraph/pull/2702)

#### ✨ New features

- Added toSet() function for removing duplicates from a list. The ordering of
  the list items is not preserved. This extension enables users to have more
  manipulation over List item objects in the query planner. Users can call the
  toSet(list) if they need to remove any duplicate values from the list.
  [#2659](https://github.com/memgraph/memgraph/pull/2659)
- Add custom authentication scheme for using OIDC Single Sign-On. Any OIDC
  scheme that exposes a public key for validating tokens using RSA is supported
  using Neo4j drivers. [#2715](https://github.com/memgraph/memgraph/pull/2715)
- Enterprise Feature: Impersonate User. Allows the user to impersonate a
  different user while executing queries. Users or roles with the correct
  permissions can define the target user for a session and execute queries as
  if the target user was logged in.
  [#2742](https://github.com/memgraph/memgraph/pull/2742)
- `project()` can now project a subgraph from lists of nodes and relationships,
  rather than just from paths. This relaxes the required format of query
  results needed to create subgraphs for further processing, or to pass to
  query modules. Users can build a subgraph from nodes and relationships which
  are computed separately, or come from different parts of your query, making
  queries cleaner, more performant, and easier to maintain.
  [#2611](https://github.com/memgraph/memgraph/pull/2611)
- Added mgp_list_reserve (C API) and List::Reserve (C++ API). Allows for more
  memory efficient list population under query modules.
  [#2650](https://github.com/memgraph/memgraph/pull/2650)
- Added query `SHOW VECTOR INDEX INFO` to display information about vector
  indices in a Cypher-like way. Although there is also a `CALL
  vector_search.show_index_info() YIELD * RETURN *`, the query was added to
  increase simplicity. The information can not be added to the `SHOW INDEX
  INFO` query because vector indices have additional information like
  dimension, size, capacity, and metric, so a different organization of records
  was needed. [#2757](https://github.com/memgraph/memgraph/pull/2757)
- Added functionality to assign multiple dynamic labels for a node from a
  variable or parameter. Users can now put dynamic label parameters in maps and
  assign labels like in the example `WITH {my_labels: ["Label1", "Label2"]} as
  x CREATE (n:x.my_labels). The result of the query is a node with 2 labels
  assigned from the list.`
  [#2713](https://github.com/memgraph/memgraph/pull/2713)
- Added support for dynamic relationship creation. Users can now create
  relationship types based on variable values. Example: `WITH {my_edge_type:
  "KNOWS"} as x CREATE ()-[:x.my_edge_type]->();` The result of this query
  would be a created triplet with the relationship type `KNOWS`.
  [#2558](https://github.com/memgraph/memgraph/pull/2558)
- Added the `length` function, which outputs the number of relationships in a
  given path object in the query or the number of elements in a list. The
  `length` function is identical to the already existing `size` function
  [#2681](https://github.com/memgraph/memgraph/pull/2681)

#### 🛠️ Improvements

- Optimise `EdgeAccessor::IsVisible` such that connectivity check is only done
  when --storage-properties-on-edges=false. This gives better performance over
  queries involving edges.
  [#2769](https://github.com/memgraph/memgraph/pull/2769)
- Added fast retrieval of approximate vertex and edge counts in the C API with
  `mgp_graph_approximate_vertex_count` and `mgp_graph_approximate_edge_count`.
  Users can now use these functions to optimize certain data structures more
  efficiently. For example, they can use this information to preallocate memory
  for `std::vector`. [#2762](https://github.com/memgraph/memgraph/pull/2762)
- The timeout is added to the SnapshotRpc message. A replica will now update
  the main on its progress when applying the snapshot received as part of the
  recovery process. [#2701](https://github.com/memgraph/memgraph/pull/2701)
- The timeout is added to the following RPC messages: AppendDeltasRpc,
  WalFilesRpc and CurrentWalRpc. The replica will now, for every 100k processed
  deltas, send the RPC message saying that the progress is made. Main will wait
  for at most 30s for the replica reply, after which it will drop the
  connection. [#2690](https://github.com/memgraph/memgraph/pull/2690)
- SHOW REPLICAS query uses a best estimation of the REPLICA's state instead of
  querying the REPLICA for its current state. The SHOW REPLICAS query now
  returns immediately, as it does not need to wait for the REPLICA response.
  This means the query is lightweight, and users can query it more frequently.
  [#2679](https://github.com/memgraph/memgraph/pull/2679)
- Improved error messages that get sent to the user in cases of query failure.
  Better understanding of the issue that caused a query to fail and easier
  debugging. Better error messages also improve final results when in an AI
  pipeline. [#2379](https://github.com/memgraph/memgraph/pull/2379)
- SHOW/USE DATABASE reclassified as a non-system query. Multiple such queries
  can be executed in parallel. Since the queries do not require unique access
  anymore, users can execute them much more frequently and without impacting
  performance. [#2674](https://github.com/memgraph/memgraph/pull/2674)
- Add diagnostic flag `--debug-query-plans` to help see all generated plans and
  their costs. Fixed filter's cost estimate and other small cost estimator
  changes. Performance improvement for index iteration. Together, they help
  provide better plans/diagnostics around plans/performance on using indexes.
  [#2722](https://github.com/memgraph/memgraph/pull/2722)
- Performance improvement around calling of query modules, less work is
  performed setting up and moving around values.
  [#2640](https://github.com/memgraph/memgraph/pull/2640)
- Performance fix. Reduced contention by ensuring batch payloads were send to
  sockets. Performance is now better when returning many small rows.
  [#2667](https://github.com/memgraph/memgraph/pull/2667)
- SystemRecoveryRPC message will now be configured with a 5min timeout.
  Recovering 1000 tenants on replica takes approximately 25s so the 5min
  timeout should cover all cases successfully. Users shouldn't observe a
  different behavior compared to the before.
  [#2654](https://github.com/memgraph/memgraph/pull/2654)

#### 🐞 Bug fixes

- When a failover is done on the replica which doesn't contain all commits as
  the old main, ForceResetRpc won't be sent anymore to the old main (now
  replica). Instead, the new main will handle the old main (now replica)'s
  history through recovery. The old main's durability files will be saved
  inside the .old directory (data_directory/snapshots and data_directory/wals)
  so that admins can always recover the newer state if needed.
  [#2727](https://github.com/memgraph/memgraph/pull/2727)
- When started in a single-tenant mode, hidden files won't be symlinked in a
  default database directory anymore. This should enable people to use Memgraph
  in K8s with various storage systems under the hood.
  [#2748](https://github.com/memgraph/memgraph/pull/2748)
- Vector index will now correctly index only vertices with the desired label
  for indexing in the dataset where there exist vertices with different labels
  but with the shared property used for indexing.
  [#2701](https://github.com/memgraph/memgraph/pull/2701)
- Coordinators will not deadlock anymore when the user requests unregistration
  of data instances. At the moment of unregistration, the cluster now needs to
  have alive main instance so that the success of the requested unregistration
  operation can be guaranteed.
  [#2735](https://github.com/memgraph/memgraph/pull/2735)
- Fixed a segmentation fault that could occur in certain scenarios when parsing
  the vector index configuration map during vector index creation.
  [#2720](https://github.com/memgraph/memgraph/pull/2720)
- Misaligned snapshot and wal timestamps cause data loss. Replication via
  snapshot and wals could lead to data loss due to misaligned timestamps.
  [#2697](https://github.com/memgraph/memgraph/pull/2697)
- Queries that access storage can now timeout if storage access cannot be
  granted. Fixes potential deadlock under heavy load. Users could see a new
  error, the access timeout. In such cases, they can retry later on or make
  sure there are no conflicting queries.
  [#2561](https://github.com/memgraph/memgraph/pull/2651)
- The action for unregistering the replication instance wasn't durable before,
  so after restarting the coordinator instance, the replication instance would
  be observed again. [#2673](https://github.com/memgraph/memgraph/pull/2673)
- Parallel recovery didn't work with WAL files. Index and constraints recovery
  would use data batches calculated from the snapshot data. This would lead to
  a failure if the snapshot is not present or partial recovery if the snapshot
  and WALs were present. Users can now recover all data correctly and in
  parallel. [#2658](https://github.com/memgraph/memgraph/pull/2658)
- If finding steps needed to recover a replica fail, the database won't be
  crashed. The replica will remain in MAYBE_BEHIND state and get recovered
  later. [#2639](https://github.com/memgraph/memgraph/pull/2639)
- Resolved issues with reloading `numpy` and `dgl` modules, which could cause
  unexpected behavior when loaded multiple times. This fix ensures
  compatibility with the latest versions of these packages.
  [#2638](https://github.com/memgraph/memgraph/pull/2638)
- Query planner fixed to handle matching multiple paths whose symbols were
  already matched. This avoids a crash on queries of the type `MATCH (n0:N)
  OPTIONAL MATCH p0 = (n0), p1 = (n0) RETURN *;`, i.e., those with multiple
  paths using the same already matched variable, such as `n0` in the given
  example. Such queries can now be run safely.
  [#2692](https://github.com/memgraph/memgraph/pull/2692)
- Query modules can now handle returning error message even when we have hit
  memory limits. We no longer terminate when trying to copying a small string
  during mgp_func_result_set_error_msg. This means we are now more robust which
  using `--memory-limit`.
  [#2702](https://github.com/memgraph/memgraph/pull/2702)
- Bugfix for subtle data race around concurrent find+remove, there was a rare
  possibility of data integrity failure that would allow multiple unique
  entries. [#2686](https://github.com/memgraph/memgraph/pull/2686)
- Fix memory leak in C++ API and in `convert.str2object`.
  [#2643](https://github.com/memgraph/memgraph/pull/2643)
- The action for unregistering the replication instance wasn't durable, so
  after restarting the coordinator instance, the replication instance would be
  observed again. [#2673](https://github.com/memgraph/memgraph/pull/2673)
- Added yielding to reader-writer locks on ARM64. This improves efficiency and
  power consumption on contented locks on machines running on the ARM64
  chipset. Users can make better use of processing resources with highly
  contented databases running on such machines.
  [#2666](https://github.com/memgraph/memgraph/pull/2666)
- Fix occasional system flakiness caused by overly optimistic memory ordering
  in `RWSpinLock`. On machines with a weak memory model (such as ARM64), this
  may occasionally have caused reads of stale data between threads. Users can
  now expect the same correct behavior consistently on all architectures.
  [#2668](https://github.com/memgraph/memgraph/pull/2668)
- Temporal functions `date()`, `localtime()`, `localdatetime()` and
  `datetime()` now accept and convert any compatible time formats. Users can
  create queries that compare different temporal types and convert between
  temporal formats. [#2630](https://github.com/memgraph/memgraph/pull/2630)
- ForceResetStorageRPC message is now configured with the timeout of 60''.
  SystemHeartbeatRPC message is now configured with the timeout of 10''. If a
  replica cannot read epoch id, the database won't crash anymore. Users
  shouldn't see any different behavior compared to the before.
  [#2648](https://github.com/memgraph/memgraph/pull/2648)
- Rewritten query planning when using Edge indexing inside Optional. Query plan
  will prefer to expand from existing nodes rather than use a new edge index to
  scan over new relationships
  [#2741](https://github.com/memgraph/memgraph/pull/2741)
- Type checking added for vector search module. There was a bug where a user
  could provide any value, which would be translated into a list of zero
  values. Now, the supported type is a list of integer and double values. Not
  providing this value will result in an exception
  [#2759](https://github.com/memgraph/memgraph/pull/2759)

### MAGE v3.1.0 - Mar 12th, 2025

#### ✨ New features

- Added migration from Amazon S3 CSV files. Users can now perform `CALL
  migrate.s3() YIELD row` syntax in order to populate their data directly from
  S3 [#547](https://github.com/memgraph/mage/pull/547)

#### 🛠️ Improvements

- The base Docker image has been switched from Debian 11 to Ubuntu 24.04 to
  ensure compatibility with toolchain-v6 and long-term security support.
  [535](https://github.com/memgraph/mage/pull/535)
- Improve static PageRank algorithm. It is now roughly 3 times faster, and it
  uses 10%-15% less memory. Additionally, it now contains an argument to set
  the number of threads used. [#552](https://github.com/memgraph/mage/pull/552)
- Improved Louvain community detection performance and reduced memory usage. A
  configuration option was added to control the number of threads used in the
  algorithm's parallelized sections. Introduced support for aborting the
  algorithm when Memgraph terminates the query using the module.
  [#549](https://github.com/memgraph/mage/pull/549)

### Lab v3.1.0 - Mar 12th, 2025

### Memgraph v3.0.0 - Jan 29th, 2025

#### ⚠️ Breaking changes

- Vector search has been removed from the experimental feature set. The API has
  changed and it's stable now.
  [#2500](https://github.com/memgraph/memgraph/pull/2500)
- Query modules memory allocator changed. Fixing an issue where query modules
  would incorrectly take up all memory due to the constant creation and
  deletion of objects. Add a reference type to CPP API, which should increase
  performance in some cases, and make minor changes to CPP API.
  [#2561](https://github.com/memgraph/memgraph/pull/2561)
- Always use durable storage under the high availability setup for coordinator
  data. From now on, HA will always use durable storage for all NuRaft data.
  The `--ha-durability` flag has been removed.
  [#2603](https://github.com/memgraph/memgraph/pull/2603)
- Remove `EdgeSetFrom`, `EdgeSetTo`, `EdgeChangeType` from the query modules
  API. These have been deemed low value and add extra complexity to other parts
  of the code base. It is essential that an edge during its lifetime is only
  ever owned by the same vertices as when the edge was created. This is
  important for the correct handling of edges. Among other things, indexing and
  edge deletion require that the edge's type, source and destination are
  invariant. [#2530](https://github.com/memgraph/memgraph/pull/2530)
- Fixed client-side routing in Kubernetes high availability deployment
  scenario. Memgraph now correctly supports client-side routing inside the
  Kubernetes environment.
  [#2568](https://github.com/memgraph/memgraph/pull/2568)

#### ✨ New features

- Added support for the exponentiation operator (^) in queries. Now, you can go
  even crazier with your math expressions under Cypher.
  [#2597](https://github.com/memgraph/memgraph/pull/2597)
- Added Cypher `toStringOrNull` function. It's practical because some data
  types are not working with `toString`. So instead of evaluating if `toString`
  could be used on a given type, it's possible to just use `toStringOrNull`
  function. [#2587](https://github.com/memgraph/memgraph/pull/2587)
- Added SHOW LICENSE INFO query to display a more granular view of what's the
  status with the license.
  [#2555](https://github.com/memgraph/memgraph/pull/2555)

#### 🛠️ Improvements

- Dynamic graph algorithms are improved and moved under the Enterprise
  offering. [#2629](https://github.com/memgraph/memgraph/pull/2629)
- Implemented slight performance improvement under the Bolt network stack by
  only using poll if timeout is needed.
  [#2632](https://github.com/memgraph/memgraph/pull/2632)
- Implemented many improvements under the replication recovery process.
  [#2617](https://github.com/memgraph/memgraph/pull/2617)
- Updated Python libraries under the Docker package due to security reasons.
  The new versions
  are: cryptography 44.0.0, pyJWT 2.10.1 and setuptools to 75.8.0.
  [#2628](https://github.com/memgraph/memgraph/pull/2628)
- Added `Clear()` function to text index for the `DROP GRAPH` command ensuring
  that when the `DROP GRAPH;` command is executed, the associated text index is
  also cleared. [#2604](https://github.com/memgraph/memgraph/pull/2604)
- Implemented various performance improvements under the `Abort` code path. You
  can expect faster abort times.
  [#2589](https://github.com/memgraph/memgraph/pull/2589)
- Vector search is no longer experimental feature, so the
  `--experimental-enabled` and `--experimental-config` should no longer be used
  with vector indices. In other words, vector search has a stable API now. You
  can now create vector index with `CREATE VECTOR INDEX index_name ON
  :Label(property) WITH CONFIG configmap;` query and drop it with `DROP VECTOR
  INDEX index_name;` query. To learn more about vector index, refer to the
  [docs page](https://memgraph.com/docs/querying/vector-search).
  [#2500](https://github.com/memgraph/memgraph/pull/2500)
- Add RPC timeouts infrastructure increasing the robustness and reliability of
  the replication and high-availability automatic failover.
  [#2607](https://github.com/memgraph/memgraph/pull/2607)

#### 🐞 Bug fixes

- Fixed a bug under the query memory tracker, causing inefficient recursive
  allocations. [#2634](https://github.com/memgraph/memgraph/pull/2634)
- Patch 3rd party dependency (RocksDB) to avoid mishandling of an exception.
  [#2618](https://github.com/memgraph/memgraph/pull/2618)
- Fixed memory ordering of the async timer because it was causing the system
  flakiness. [#2625](https://github.com/memgraph/memgraph/pull/2625)
- Fixed the `PeriodicCommit` crash if the commit generates replication error.
  [#2598](https://github.com/memgraph/memgraph/pull/2598)
- Corrected logic for removing edge type indices entries during an abort. This
  fix stops segfaults that could occur non-deterministically.
  [#2596](https://github.com/memgraph/memgraph/pull/2596)
- Fixed deadlock by releasing RPC lock before acquiring the engine lock. When
  checking replica state, main will release RPC lock before taking the engine
  lock to avoid deadlock with commit thread.
  [#2515](https://github.com/memgraph/memgraph/pull/2515)
- Fixed deadlock when reading replication state during commit.
  [#2577](https://github.com/memgraph/memgraph/pull/2577)
- Fixed data race with edge type indexes and their cleanup. This fix prevents
  potential for segfaults.
  [#2588](https://github.com/memgraph/memgraph/pull/2588)
- Fixed bug of delta objects not releasing after unique constraint violation.
  Unique constraint error would release the deltas to GC as part of Abort, but
  not make the commit timestamp as finished. Hence the deltas and any future
  deltas would be prevented from being deleted during GC resulting in a memory
  leaks. [#2566](https://github.com/memgraph/memgraph/pull/2566)
- Fixed perf regression by moving the GC work to be done less frequently.
  Recent fix for ensuring memory was being released caused a noticeable cost to
  throughput for single_vertex_write benchmark.
  [#2576](https://github.com/memgraph/memgraph/pull/2576)
- Fix `convert.str2object` in C++ to work for primitives.
  [#2559](https://github.com/memgraph/memgraph/pull/2559)
- Fix MATCH after CALL not being able to execute a valid plan. It's possible
  now to use MATCH after CALL clause.
  [#2550](https://github.com/memgraph/memgraph/pull/2550)

### MAGE v3.0.0 - Jan 29th, 2025

#### ⚠️ Breaking changes

- uuid_generator now uses a different method of generating UUIDs so it might
  not be compatible with the one in previous versions.
  [#538](https://github.com/memgraph/mage/pull/538)
- Rewrite refactor modules not to use deleted API (`EdgeSetTo`, `EdgeSetFrom`
  and `EdgeChangeType` were removed from Memgraph under
  [#2530](https://github.com/memgraph/memgraph/pull/2530)).
  `refactor.from`, `refactor.to`, `refactor.invert`, `refactor.delete_and_reconnect`,
  and `refactor.rename_type` now create a new edge instead of modifying the old
  one. This requires changes if you rely on an internal edge ID (previously, it
  would stay the same; now, it changes due to the creation of a new edge).
  [#536](https://github.com/memgraph/mage/pull/536)
- Dynamic algorithms were removed from MAGE and moved under Memgraph
  Enterprise. [#540](https://github.com/memgraph/mage/pull/540)

#### 🛠️ Improvements

- Rewrite the UUID module to be the same as on the
  [memgraph/memgraph](https://github.com/memgraph/memgraph) repo.
  [#538](https://github.com/memgraph/mage/pull/538)

#### 🐞 Bug fixes

- Fixed DGL issues. [#540](https://github.com/memgraph/mage/pull/540)

### Lab v3.0.1 - Jan 30th, 2025

#### 🐞 Bug fixes

- Fixed node icons in query data results and
  graph schema, ensuring correct visualization
  and consistency.

### Lab v3.0.0 - Jan 29th, 2025

#### ⚠️ Breaking changes

- **GraphChat Migration**: Messages and LLM setups from
  previous versions of the Lab will not be migrated to
  the new GraphChat due to the complexity and enhanced
  organizational structure introduced in this update.

#### ✨ New features

- Organize your GraphChat conversations with
  the ability to create multiple message threads,
  improving structure and clarity.
- Set up DeepSeek LLM models for your Graph Chat,
  expanding your model options and customization.
- Instead of being limited to a single model per
  provider (e.g., Ollama, Azure OpenAI, OpenAI,
  DeepSeek) in your Graph Chat, you can now create
  and manage multiple models for each provider, each
  with unique configurations and model references.
- GraphChat conversation history is now used as a
  reference, making chats contextually aware.
  Configure the size of the conversation context
  as needed.
- Add feedback to individual messages in GraphChat
  to exclude them from future context, improving
  conversation precision.
- Leverage multiple GraphChat views within Lab
  layouts to run LLM questions in parallel.
- Each GraphChat answer includes step-by-step details,
  queries, token counts, and schema references,
  simplifying debugging and analysis.
- GraphChat will automatically retry Cypher generation
  for failed queries. Configure retry limits and view
  details for each attempt.
- Enterprise license users can now [manage Graph Style Script
  (GSS)](https://memgraph.com/docs/data-visualization/user-manual/graph-style-script-sharing) with
  [remote storage](https://memgraph.com/docs/data-visualization/user-manual/remote-storage), enabling
  team collaboration and script sharing. You can balance between locally stored
  and remotely saved scripts to suit your workflow.
- A dismissable notice now alerts you if `vm.max_map_count`
  is not configured correctly for [optimal Memgraph performance](https://memgraph.com/docs/database-management/system-configuration#increasing-memory-map-areas).
- Switch Lab to dark mode from the application settings.

#### 🛠️ Improvements

- The graph schema view now displays property count, fill
  percentages and property types for all nodes and
  relationships in the database. To access full graph
  information, ensure Memgraph is running with the
  `--schema-info-enabled` [configuration](https://memgraph.com/docs/querying/schema#run-time-schema-tracking).
- When connecting to a coordinator instance in a Memgraph
  HA environment (version 3.0.0 or later), the status bar
  now displays real-time information about whether you're
  connected to a leader or a follower node.

#### 🐞 Bug fixes

- Screenshots are now generated as expected when saving
  graph style scripts from query execution.

### Memgraph v2.22.1 - Dec 22, 2024

#### 🛠️ Improvements

- The `convert.str2object` function was rewritten from Python to C++ so Python
  GIL does not bottleneck it. Users are now able to achieve the necessary
  parallelization improvements.
  [#2532](https://github.com/memgraph/memgraph/pull/2532)
- Added `SHOW INSTANCE` query to expose more information about the coordinator
  instance to which the user is connected. The query returns: `instance_name`,
  `bolt_server`, `coordinator_server`, `management_server` and `cluster_role`.
  [#2526](https://github.com/memgraph/memgraph/pull/2526)
- Added `REMOVE COORDINATOR <ID>` query. Only followers can be removed from
  Raft logs. The request to remove a leader won't be accepted in Raft logs.
  [#2533](https://github.com/memgraph/memgraph/pull/2533)
- Updated `librdkafka` to v2.6.1. You can expect better stability on the Kafka
  streaming side. [#2523](https://github.com/memgraph/memgraph/pull/2523)

#### 🐞 Bug fixes

- Fixed a bug when a replica received an old WAL file by ignoring it.
  Previously, a replica could receive a WAL file that contains all old data. If
  that is the case, the old epoch would be added to the history together with
  the current durable timestamp on the replica, causing the wrong information
  contained inside the replica. Replicas shouldn't naively use data from WAL
  files received. [#2552](https://github.com/memgraph/memgraph/pull/2552)
- Fixed a subtle memory leak. The skip_list used for async timers was
  previously not being collected. `--query-execution-timeout-sec` caused an
  allocation that was small but not free.
  [#2539](https://github.com/memgraph/memgraph/pull/2539)
- Fixed error message when unloading query message with a correct description
  of what happens. An additional reason why the unloading of a module might
  fail is that the module might not exist.
  [#2547](https://github.com/memgraph/memgraph/pull/2547)
- Fixed planner so it can always generate at least one valid plan. Valid plans
  are currently all the ones that start expanding from a node. There was an
  edge case where no valid plan could be constructed, as all constructed plans
  would start expanding from an edge without edge indices in place.
  [#2543](https://github.com/memgraph/memgraph/pull/2543)
- Fixed the incorrect management of the delta counter metric after the abort.
  `SHOW STORAGE INFO` under `unreleased_delta_objects` should be better now.
  [#2544](https://github.com/memgraph/memgraph/pull/2544)
- Fixed the case when the database would crash without any valid query plan
  generated for a given query. Instead of the crash, the database will throw
  the `QueryException`. [#2537](https://github.com/memgraph/memgraph/pull/2537)
  and [#2535](https://github.com/memgraph/memgraph/pull/2535)
- Fixed issue with list pattern pattern comprehension under arbitrary
  expression. List pattern comprehension is only allowed under `WITH` and
  `RETURN` clauses. [#2534](https://github.com/memgraph/memgraph/pull/2534)
- Fixed a bug under after-commit triggers where the database accessor becomes
  invalid after the transaction finishes. Before the fix, Memgraph would crash
  if the trigger used the access to access the changed entities under the
    graph. [#2536](https://github.com/memgraph/memgraph/pull/2536)
- Fixed the case when recovery and promotion are executed concurrently on a
  replica. The promotion request will wait for recovery to finish before
  preparing storage for the new epoch.
  [#2521](https://github.com/memgraph/memgraph/pull/2521)
- Fixed how `metrics` keyword was added to the grammar. The keyword couldn't be
  used as a variable name.
  [#2502](https://github.com/memgraph/memgraph/pull/2502)

### MAGE v1.22.1 - Dec 22, 2024

- The MAGE version was only bumped because there is a new version of Memgraph
  (v2.22.1). Nothing changed under the MAGE itself.

### Memgraph v2.22.0 - Nov 27, 2024

#### ⚠️ Breaking changes

- `--experimental-enabled=high-availability` does not exist anymore. If you
  used the feature, please remove the experimental flag because it's no longer
  necessary. [#2479](https://github.com/memgraph/memgraph/pull/2479)
- `any(v IN [] WHERE v)` was returning `Null`; since this release, it returns
  `False`. [#2489](https://github.com/memgraph/memgraph/pull/2489)
- We improved global settings error messages (log to stderr and log level);
  this might be a breaking change if you rely on exact error message text.
  [#2494](https://github.com/memgraph/memgraph/pull/2494)

#### ✨ New features

- The first version of the vector search feature was added. To enable and
  configure the feature, use `--experimental-enabled=vector-search` and
  `--experimental-config` flags. Learn more under the [vector search
  documentation page](https://memgraph.com/docs/querying/vector-search).
  [#2406](https://github.com/memgraph/memgraph/pull/2406)
- The data snapshotting (point-in-time backups of the graphs) management has
  been improved. It's possible now to use crontab syntax to specify when the
  snapshot will be created.
  [#2435](https://github.com/memgraph/memgraph/pull/2435)
- Added `IsEmpty` function that accepts `Null` (`IsEmpty(Null)` returns
  `Null`), `List`, `Map`, `String` and returns `Boolean` value.
  [#2489](https://github.com/memgraph/memgraph/pull/2489)

#### 🛠️ Improvements

- High-availability automatic failover became the stable feature 🎉
  `--experimental-enabled=high-availability` flag is not required anymore.
  Please visit [high-availability](https://memgraph.com/docs/clustering/high-availability) page to learn
  more about the feature.
  [#2479](https://github.com/memgraph/memgraph/pull/2479)
- The planner now utilizes the point index for `point.withinbbox` queries and
  makes `point.distance` queries using the index without the `WITH` statement.
  This improvement results in faster queries when using the mentioned spatial
  features. [#2449](https://github.com/memgraph/memgraph/pull/2449)

#### 🐞 Bug fixes

- Hiding passwords under logs wasn't handled well in case of double quotes.
  That's fixed now, so you don't have to worry that the passwords will get
  exposed. [#2488](https://github.com/memgraph/memgraph/pull/2488)
- Fix crashes (only in debug mode) when using point functions, for example, by
  providing x, y, z, and CRS for 2D points and vice versa.
  [#2463](https://github.com/memgraph/memgraph/pull/2463)
<details>
<summary>
  High-availability automatic failover updates
</summary>
<span>
  The code changes below improve the high-availability automatic failover,
  resulting in a more stable Memgraph runtime in that environment.
</span>
<ul>
  <li>
    - Properly handle failures in WAL replication. An interpreter will now throw an
    exception during commit instead of calling abort when the main becomes a
    replica with a write transaction to be committed. Furthermore, the replica
    instance won't crash if it receives an existing vertex.
    <a href="https://github.com/memgraph/memgraph/pull/2467">[#2467]</a>
  </li>
  <li>
    - The system will now correctly handle situations where the main gets re-elected as
    main again. <a href="https://github.com/memgraph/memgraph/pull/2472">[#2472]</a>
  </li>
  <li>
    - The `SHOW INSTANCES` query will now run correctly when the leader becomes a
    follower. <a href="https://github.com/memgraph/memgraph/pull/2474">[#2474]</a>
  </li>
  <li>
    - Fix that will not allow a sync replica to be READY while being behind by
  aligning the recovery timestamp check with timestamp modifications. <a href="https://github.com/memgraph/memgraph/pull/2476">[#2476]</a>
  </li>
  <li>
    - When a main instance misses a failover and is selected to be the new main, Raft
    generates a new UUID. Previously, replicas wouldn't update their durability,
    and MAIN wouldn't update its clients. Now, when swapping UUIDs,
    durability and active clients are updated correctly.
    <a href="https://github.com/memgraph/memgraph/pull/2478">[#2478]</a>
  </li>
  <li>
    - Users must set `--coordinator-hostname` when starting a coordinator in HA.
    Previously, the database would crash; now, it logs an error message.
    <a href="https://github.com/memgraph/memgraph/pull/2482">[#2482]</a>
  </li>
  <li>
    - Instance-level clients can now safely be closed after committing. Previously, a
    promotion request executed in parallel with committing could cause data instance locking.
    <a href="https://github.com/memgraph/memgraph/pull/2485">[#2485]</a>
  </li>
</ul>
</details>

### MAGE v1.22.0 - Nov 27, 2024

#### 🐞 Bug fixes

- [Link
  prediction](https://memgraph.com/docs/advanced-algorithms/available-algorithms/gnn_link_prediction),
  [node
  classification](https://memgraph.com/docs/advanced-algorithms/available-algorithms/gnn_node_classification)
  and
  [text](https://memgraph.com/docs/advanced-algorithms/available-algorithms/text)
  modules are available again.
  [#528](https://github.com/memgraph/mage/pull/528)
- The built-in authentication module dependencies are now pre-installed also
  under MAGE Docker images. This mainly refers to the built-in SSO modules that
  depend on third-party libraries to offer a secure way of authenticating into
  Memgraph instances. [#526](https://github.com/memgraph/mage/pull/526)

### Lab v2.19.1 - Nov 27, 2024

#### 🐞 Bug fixes

- Resolved an issue where username and password inputs were not displayed
  correctly on the login screen in the Desktop app when authentication
  was required.

### Lab v2.19.0 - Nov 27, 2024

#### ✨ New features

- The Lab web application can now be started on a [custom base
  path](https://memgraph.com/docs/data-visualization/user-manual/custom-base-path) using the environment
  variable `BASE_PATH`, which is useful for reverse proxy configurations.

#### 🛠️ Improvements

- In multi-tenant environments, a search bar is added to filter databases by
  name.
- You can now view the complete Memgraph URI and Memgraph version by clicking on
  the connection in the top left corner.

#### 🐞 Bug fixes

- Fixed an issue where logs failed to disconnect
  from the previous Memgraph connection in the Desktop
  app.
- The authentication screen for entering a username
  and password when connecting to Neo4j now displays
  correctly.

### Memgraph v2.21.0 - Nov 6, 2024

#### Upgrading to Memgraph v2.21.0 breaking change

> **Info**
>
> Coordinator log store directory in `--data-directory` needs to be started from
> scratch since new type of Raft log is added and some old ones were deleted.
> Memgraph could crash without deleting old logs from log store directory.

#### New features and improvements

- Added planner optimization to replace `point.distance` filter with an index
  scan. Queries like `WITH point({x:0,y:0}) AS a MATCH (m:L1) WHERE
  point.distance(m.prop, a) < 1 RETURN m;`, should run faster if there is the
  [point index](https://memgraph.com/docs/fundamentals/indexes#point-index) created. If you are dealing
  with spatial data, there is a high chance you need such indexes to run your
  queries faster. [#2362](https://github.com/memgraph/memgraph/pull/2362)
- Significantly improved schema retrieval performance in both TRANSACTIONAL and
  ANALYTICAL setups. Feel free to use `--schema-info-enabled=True` to get
  schema instantly by running the `SHOW SCHEMA INFO;` query.
  [#2383](https://github.com/memgraph/memgraph/pull/2383)
- Added query metadata when logging incoming query. It's possible now, e.g., to
  filter certain database log entries or search for specific items in the log.
  [#2404](https://github.com/memgraph/memgraph/pull/2404)

#### Bug fixes

- Closing instance would crash due to streams and/or TTL running in the
  background. [#2361](https://github.com/memgraph/memgraph/pull/2361)
- Add backticks around property keys in any \{key: value\} patterns.
  [#2371](https://github.com/memgraph/memgraph/pull/2371)
- When replica gets promoted to main upon failover, it will use correctly
  incremented durable timestamp.
  [#2366](https://github.com/memgraph/memgraph/pull/2366)
- `schema.assert` will throw an exception if ran inside the explicit transaction.
  [#2352](https://github.com/memgraph/memgraph/pull/2352)
- Modifying `mgp::memory` is prohibited; use `MemoryDispatcherGuard` instead.
  Under C++ query modules, if module tries to set `mgp::memory` there will be a
  compilation error. Usually at the beginning of the `mgp_init_module`
  implementation there should be the `mgp::MemoryDispatcherGuard
  guard(memory);`. Usage examples could be found under [MAGE C++
  modules](https://github.com/memgraph/mage/tree/main/cpp).
  [#2424](https://github.com/memgraph/memgraph/pull/2424)
- Fixed incorrect usage of cv and stop token in scheduler and thread pool.
  [#2452](https://github.com/memgraph/memgraph/pull/2452)
- Fixed unreleased deltas counter. There was a missing case of tracking when
  deltas were actually released. This gave the appearance in `SHOW STORAGE
  INFO;` that we were not actually releasing them.
  [#2446](https://github.com/memgraph/memgraph/pull/2446)
- Fixed edge index not removing deleted edges. Edge index would not remove
  deleted (obsolete) entries, while the GC would delete the edge object. Edge
  type index was impacted, while the property type would detect that the object
  was deleted due to it checking the property.
  [#2439](https://github.com/memgraph/memgraph/pull/2439)
- Removed usage of RTLD_DEEPBIND under query modules. Query modules, especially
  the Python ones should work better in terms of correct module loading.
  [#2347](https://github.com/memgraph/memgraph/pull/2347)

#### High-availability automatic failover improvements

There are a lot of efforts to make Memgraph HA rock-solid in the production
environments. Under this release, the feature is still experimental, but
getting very close to put the stable flag next to it. The full list if issues
and improvements can be found
[HERE](https://github.com/memgraph/memgraph/issues?q=label%3Ajepsen) (GH issues
labeled with `jepsen`).

- All data instances now use one type of callbacks to simplify state
  reconciliation. [#2421](https://github.com/memgraph/memgraph/pull/2421)
- Incomplete storage reset when MAIN force reset or applied snapshots to
  REPLICA. [#2422](https://github.com/memgraph/memgraph/pull/2422)
- Coordinators now use one type of Raft logs to remove the need for distributed
  locks. Coordinator log store needs to be started from scratch, hence this
  change is a breaking one.
  [#2430](https://github.com/memgraph/memgraph/pull/2430)
- Fixed deadlock when demoting data instance. When a coordinator sends a
  request for promotion or demotion, the lock on replication state is taken so
  that it cannot be interleaved with commit. If main became a replica and was
  executing write transaction, the transaction will be aborted. Scheduler is
  now stopped before thread pool in RPC client.
  [#2447](https://github.com/memgraph/memgraph/pull/2447)

### MAGE v1.21.0 - Nov 6, 2024

#### New features and improvements

- Added [Leiden
  algorithm](https://memgraph.com/docs/advanced-algorithms/available-algorithms/leiden_community_detection).
  The implementation is "static" in a sense that it's operating on the whole
  graph. Leiden is a modern community detection algorithm that in Memgraph's
  implementation provides hierarchies of communities. A big use-cases for
  Leiden community detection is [GraphRAG](https://memgraph.com/docs/ai-ecosystem/graph-rag).
  [#508](https://github.com/memgraph/mage/pull/508)

### Lab v2.18.2 - Nov 7, 2024

#### Bug fixes

- You can now edit or remove connections with long
  names that previously overrode the edit/remove action.
- Fixed issue where disk usage in the status bar
  showed memory usage instead (bug introduced in 2.18.0).
- Removed the "Danger Zone" from the overview.

### Lab v2.18.1 - Nov 6, 2024

#### Bug fixes

- Resolved error with newer Ollama models causing
  "Cannot read properties of undefined (reading 'match')"
  when using the GraphChat feature.

### Lab v2.18.0 - Nov 6, 2024

#### New features and improvements

- Enable HTTPS in Lab without needing a reverse proxy,
  using self-signed or custom certificates with environment
  variables `SSL_IS_ENABLED`, `SSL_CERT_PATH` and
  `SSL_KEY_PATH`.
- Logs in the Lab no longer show system Cypher calls with
  Memgraph version 2.21 or above.

#### Bug fixes

- SSO redirect for Lab when configured with HTTPS works
  as expected now.

### Memgraph v2.20.1 - Oct 10, 2024

#### Bug fixes

- When a replica gets promoted to main upon [failover](https://memgraph.com/docs/clustering/high-availability#failover), it will now use correctly incremented durable timestamp. [#2366](https://github.com/memgraph/memgraph/pull/2366)
- Replica will now take MAIN's [uuid](https://memgraph.com/docs/clustering/high-availability#how-replica-knows-which-main-to-listen) only when REPLICA's current sequence number is 0. [#2388](https://github.com/memgraph/memgraph/pull/2388)
- Eliminated deadlock caused by concurrent calls to Stop method on the Scheduler. Eliminated deadlock caused by calling Pause after Stop from different threads. [#2398](https://github.com/memgraph/memgraph/pull/2398)
- Fixed a problem where [EdgeType index](https://memgraph.com/docs/fundamentals/indexes#edge-type-index) would assign every element as obsolete which then caused the index to be incorrectly emptied by the GC. Fixed another problem where [EdgeTypeProperty index](https://memgraph.com/docs/fundamentals/indexes#edge-type-property-index) never ran its GC. Now properly prevents edge index creation when properties-on-edges is disabled - the user must explicitly [provide](https://memgraph.com/docs/database-management/configuration#storage) --storage-properties-on-edges. Now correctly prevents Edge indices being recovered when properties-on-edges is disabled by causing an assert during recovery. [Replication](https://memgraph.com/docs/clustering/replication) now updates edge indices. [#2390](https://github.com/memgraph/memgraph/pull/2390)
- Replica now correctly reflects database registration success or failure.  When replica receives a request for registering replica, it will now return status code NOT_MAIN.  If attempting to undo an incomplete registration fails, then the Memgraph instance will be intentionally crashed because the system is in unrecoverable state.  Main instance can now deal with stale RPC message GetReplicaUUID without crashing the instance.  [#2373](https://github.com/memgraph/memgraph/pull/2373)
- Closing instance no longer crashes due to [streams](https://memgraph.com/docs/data-streams) and/or [TTL](https://memgraph.com/docs/querying/time-to-live) running in the background. [#2361](https://github.com/memgraph/memgraph/pull/2361)

### Lab v2.17.0 - Sep 25, 2024

#### New features and improvements

- A monitoring dashboard is available, offering
  enhanced insights into the state of your data and potential
  optimization opportunities for your Memgraph
  Enterprise instance.
- Choose a custom model for the OpenAI GraphChat integration
  instead of being limited to the default `gpt-4`.
- GraphChat is more consistent with its responses due to a
  change in the LLM temperature setting from `0.7` to `0.01`.
- GraphChat will attempt to correct invalid Cypher queries
  up to three times, ensuring more reliable query generation.
- Graph schema generation is now almost instant when the
  Memgraph configuration parameter `schema-info-enabled` is
  set to `true`.

#### Bug fixes

- Drag-and-drop sorting of query collections now
  maintains the correct order as expected.
- User permissions, status bar stats, and replication status
  are now aligned. Instead of breaking the connection due to
  permission issues, users will now see "N/A" where applicable.

### Memgraph v2.20.0 - Sep 25, 2024

#### Upgrading to Memgraph v2.20.0 breaking change

> **Info**
>
> In this version of Memgraph, we have removed the experimental classification for
> system replication. Going forward, the `--experimental-enabled` configuration
> flag no longer recognizes the `system-replication` argument which was introduced
> in v2.15.
> [System replication](https://memgraph.com/docs/clustering/replication/system-replication) is now a fully
> supported Enterprise-licensed feature.

#### New features and improvements

- [Enterprise] Graduated replication to “production” from “experimental”. [#2295](https://github.com/memgraph/memgraph/pull/2295)
- Up to date schema tracking, available for In Memory Transactional and In Memory Analytical. [#2256](https://github.com/memgraph/memgraph/pull/2256)
- New index on geospatial datatype. [#2298](https://github.com/memgraph/memgraph/pull/2298)
- Added SASL2 support for Memgraph with Kafka streaming. [#2305](https://github.com/memgraph/memgraph/pull/2305)
- Query planner now leverages edge type indices and edge-property indices. [#2293](https://github.com/memgraph/memgraph/pull/2293)
- New session trace feature for fine grained inspection of internal engine events. [#2276](https://github.com/memgraph/memgraph/pull/2276)
- New type constraint feature. [#2289](https://github.com/memgraph/memgraph/pull/2289)

#### Bug fixes

- Reverted a performance regression related to logging. [#2316](https://github.com/memgraph/memgraph/pull/2316)
- No longer skips post_update callback when recovering flag value data from durability module. [#2274](https://github.com/memgraph/memgraph/pull/2274)
- Now correctly cleans any edge indices during abort. [#2310](https://github.com/memgraph/memgraph/pull/2310)
- No longer adds entries to the wrong label+property index. [#2294](https://github.com/memgraph/memgraph/pull/2294)
- Remove deadlock during streaming database destruction. [#2288](https://github.com/memgraph/memgraph/pull/2288)
- Fix IPv6 endpoint version resolution. [#2281](https://github.com/memgraph/memgraph/pull/2281)
- Fix failover recovery when missing WAL or invalid state machine. [#2241](https://github.com/memgraph/memgraph/pull/2241)
- Cartesian to Indexed Join was happening only for property filters. [#2343](https://github.com/memgraph/memgraph/pull/2343)
- Closed several defects in replication system. [#2283](https://github.com/memgraph/memgraph/pull/2283)
- Altered “become leader” callback HA logic to always do check of whole cluster to avoid data loss. [#2248](https://github.com/memgraph/memgraph/pull/2248)

### Memgraph v2.19.0 - Aug 14, 2024

#### Breaking changes

- Fixed an issue where `LocalDateTime` did not convert user date to UTC
properly. `LocalDateTime` now considers the set database [timezone for
conversions](https://memgraph.com/docs/fundamentals/data-types#localdatetime). A new [command-line
argument and database setting](https://memgraph.com/docs/fundamentals/data-types#database-timezone) for
timezone have been added, allowing users to define the database timezone for
accurate `LocalDateTime` conversions.
[#2204](https://github.com/memgraph/memgraph/pull/2204)

- The `--data-recovery-on-startup` flag is now set to `true` by default. This
ensures that all user data will be recovered on startup without needing to
specify any configuration options.
[#2178](https://github.com/memgraph/memgraph/pull/2178)

#### New features and improvements

- [Enterprise Ed.] Added functionality to hide sensitive information,
specifically passwords, in query logs. Passwords in queries are now masked with
asterisks (*).
[#2140](https://github.com/memgraph/memgraph/pull/2140)

- Introduced the [`SHOW ACTIVE USERS
INFO`](https://memgraph.com/docs/database-management/server-stats#active-users-information) command,
which allows users to view a list of active users along with detailed session
information.
[#2183](https://github.com/memgraph/memgraph/pull/2183)

- Added support for [periodic
commits](https://memgraph.com/docs/querying/read-and-modify-data#periodic-execution) within Cypher
queries, allowing for batched execution. This enhancement prevents memory
exhaustion by committing intermediate results during the execution of large
queries.
[#2182](https://github.com/memgraph/memgraph/pull/2182)
[#2194](https://github.com/memgraph/memgraph/pull/2194)
[#2196](https://github.com/memgraph/memgraph/pull/2196)
[#2197](https://github.com/memgraph/memgraph/pull/2197)
[#2217](https://github.com/memgraph/memgraph/pull/2217)

- Added the Time-to-Live (TTL) feature, enabling users to tag nodes with an
expiration time. A background thread automatically deletes these nodes and their
associated edges once they expire.
[#2167](https://github.com/memgraph/memgraph/pull/2167)

- Users can now define their passwords directly using a hash, allowing for more
secure authentication methods.
[#2148](https://github.com/memgraph/memgraph/pull/2148)

- Added property compression across the graph using the zlib compression
library. This feature introduces [new configuration
flags](https://memgraph.com/docs/fundamentals/storage-memory-usage#property-compression)
`--storage-property-store-compression-enabled` and
`--storage-property-store-compression-level`, allowing users to enable
compression and set the desired compression level.
[#2102](https://github.com/memgraph/memgraph/pull/2102)

#### Bug fixes

- Fixed an issue where `SHOW REPLICAS` was returning an empty map for
`data_info` in the Community Edition due to an incorrect database name check.
[#2236](https://github.com/memgraph/memgraph/pull/2236)

- Optimized the deletion process for nodes and edges by deferring the removal of
obsolete index entries to the garbage collector (GC). This change improves
deletion performance by reclaiming memory after the GC cycle instead of
immediately.
[#2242](https://github.com/memgraph/memgraph/pull/2242)

- Fixed an issue where coordinators did not yield leadership according to the
`leadership_expiry_` flag. The leadership expiry is now correctly respected,
preventing the use of
`heart_beat_interval_*raft_server::raft_limits_.response_limit_` as the expiry
limit for nodes deemed unhealthy.
[#2224](https://github.com/memgraph/memgraph/pull/2224)

- Leadership now expires as it was intended, before the election period starts.
[#2224](https://github.com/memgraph/memgraph/pull/2224)

- Fixed a bug where no default database was selected when `MG_ENTERPRISE=OFF`.
[#2267](https://github.com/memgraph/memgraph/pull/2267)

- Fixed the endpoint for session closure by a peer and ensured it is correctly
reflected in the audit logs.
[#2181](https://github.com/memgraph/memgraph/pull/2181)

- Incorrect cache of query plan resulted in an incorrect sharing of resources
for some call procedures, which in rare cases could cause segfault. Regression
was introduced in v2.18.1.
[#2268](https://github.com/memgraph/memgraph/pull/2268)

### MAGE v1.19 - Aug 14, 2024

#### Bug fixes

- [Node2Vec](https://memgraph.com/docs/advanced-algorithms/available-algorithms/node2vec) now requires
either `hs` set to 1 or `negative` set to a positive integer instead of a
non-negative integer due to bumping Gensim version.
[#502](https://github.com/memgraph/mage/pull/502)

### Lab v2.16.0 - Aug 14, 2024

#### New features and improvements

- Select [different connection types](https://memgraph.com/docs/data-visualization/user-manual/connection-types):
  standalone Memgraph, Memgraph in replication mode, and Memgraph in HA
  cluster mode.
- Connect to the Memgraph HA cluster and see the cluster status of your
  coordinators and data instances in real-time.
- Connect to the Memgraph main instance in replicated mode and see the
  replication status in real-time.
- Connect directly to the Coordinator to run cluster management queries.
- Configure optional proxy endpoint for the OpenAI model in the GraphChat.
- Choose a custom model for the Ollama GraphChat integration instead of
  the default model `llama2`.
- Graph results on retina displays have 15% faster rendering with better
  display quality.
- For empty or invalid Graph Style Script you will receive a warning
  message instead of empty graph results.
- Clearing a database before loading a predefined dataset will use less
  memory now because it removes data in batches.
- Clearing a database before loading a predefined dataset will remove
  triggers too along with nodes, relationships, indexes and constraints.
- Set environment variable `LOG_STACKTRACE_IS_ENABLED` to true
  to see the stack trace for the error logs.

#### Bug fixes

- Display graph options are now applied correctly if you pick a color/text
  before choosing the option "This node" or "All nodes with the same label".

### Memgraph v2.18.1 - Jul 23, 2024

#### Breaking changes

- Introduced [management servers on coordinators](https://memgraph.com/docs/clustering/high-availability)
to get the health state of the cluster. Each coordinator now has a management
server and client connection to all other coordinators. This allows follower
coordinators to correctly report health states by contacting the leader
coordinator. [#2138](https://github.com/memgraph/memgraph/pull/2138)

#### New features and improvements

- Added current amount of unreleased delta objects to the [`SHOW STORAGE
INFO`](https://memgraph.com/docs/configuration/server-stats) query.
[#2066](https://github.com/memgraph/memgraph/pull/2066)

- Added a new query, [`SHOW METRICS
INFO`](https://memgraph.com/docs/database-management/server-stats#metrics-information), which exposes the
same metrics available from the HTTP endpoint, now accessible via a query.
[#2052](https://github.com/memgraph/memgraph/pull/2052)

#### Bug fixes

- Fixed an issue with the wrong handling of NuRaft config logs in HA durability,
where the logs were parsed from RocksDB as text although they are serialized in
NuRaft binary format. [#2189](https://github.com/memgraph/memgraph/pull/2189)

- Fixed edge type property index lookup scan. The edge type property index scan
now correctly fills vertices in the frame, preventing queries that construct a
path based on edge type property index from returning nulls. The edge type
property index scan is disabled for more complex queries, defaulting to
`ScanAll` instead.
[#2190](https://github.com/memgraph/memgraph/pull/2190)

- Controlled delta allocations to improve memory management in
`IN_MEMORY_TRANSACTIONAL` mode. Deltas now use a dedicated PMR allocator for
property values and other delta allocations, allowing for quick and trivial
destruction of deltas post-query. This change aligns resident memory with
expectations for `IN_MEMORY_ANALYTICAL`, which never allocates deltas.
[#2151](https://github.com/memgraph/memgraph/pull/2151)

- In HA, the election timeout is now set between 2000ms and 4000ms, with the
heartbeat interval occurring every second. These values are chosen to tolerate
occasional network hiccups without triggering unnecessary leadership changes.
[#2207](https://github.com/memgraph/memgraph/pull/2207)

- C and C++ query modules are a lot faster now and no longer slow down when ran
concurrently.
[#2176](https://github.com/memgraph/memgraph/pull/2176)

- Correctly handle memory limits when triggered by allocations in third-party
code.
[#2191](https://github.com/memgraph/memgraph/pull/2191)

- Fixed the enums dump function to correctly print all enum values.
[#2168](https://github.com/memgraph/memgraph/pull/2168)

- Correction for `propertySize` function when enums are inside Lists/Maps. The
internal reader now reads the correct number of bytes, preventing issues where
the reader was positioned incorrectly.
[#2205](https://github.com/memgraph/memgraph/pull/2205)

### Memgraph v2.18.0 - Jul 3, 2024

#### Upgrading to Memgraph v2.18.0

> **Info**
>
> If you are upgrading to Memgraph v2.18.0 from any previous versions and you
> have a persistent configuration file (docker volumes or native install) due to
> the [breaking changes in the configuration
> file](https://memgraph.com/docs/release-notes#memgraph-v2180---jul-3-2024) you need to set manually
> `--data-recovery-on-startup=true` to restore data on startup.
>
> If you do not set `--data-recovery-on-startup=true`, Memgraph will start without
> data recovery and your data (snapshots at `var/lib/memgraph/snapshots`) will be
> overwritten by the newly created periodical snapshots, potentially causing data
> loss, depending on the configurations of snapshot creation.
>
> If you think you have experienced data loss and added
> `--data-recovery-on-startup=true`, stopping Memgraph creates a new snapshot by
> default, and starting Memgraph will restore the data from the last snapshot,
> there is a possibility that the snapshot with data is still present in the
> `var/lib/memgraph/snapshots` folder, depending on how long Memgraph was running
> and how many times it was restarted.
>
> To prevent data loss in all cases, before upgrading to Memgraph v2.18.0,
> create a
> [backup of your data directory](https://memgraph.com/docs/deployment/linux#backup-data) (`var/lib/memgraph`).

#### Breaking changes

- Configuration flag `--coordinator-hostname` is added, which is used to fix the
issue of having `0.0.0.0` in the output of `SHOW INSTANCES` on the first leader.
The flag has its environment variable equivalent in
`MEMGRAPH_COORDINATOR_HOSTNAME`. Now all coordinator instances [need to provide
a configuration option](https://memgraph.com/docs/clustering/high-availability) for the coordinator
hostname. For local deployment, use `localhost`. For containerized deployments,
use the name of the container. This change is breaking because coordinators
cannot be started if this configuration option isn't set.
[#2131](https://github.com/memgraph/memgraph/pull/2131)

- The Coordinator's durability has been reinforced through enhanced storage of
RAFT's logs and snapshots, now persistently stored in the
`coordinator/raft_data/logs` folder. The [coordinator
configuration](https://memgraph.com/docs/clustering/high-availability) is securely maintained in the
`coordinator/raft_data/network` folder, which is kept constantly durable. A new
flag, `--ha-durability`, allows for optional deactivation of this log and
snapshot durability if required.
[#2029](https://github.com/memgraph/memgraph/pull/2029)

- Removed [deprecated command-line flags](https://memgraph.com/docs/database-management/configuration) to
  streamline configuration and improve system performance. The following flags
  are no longer supported: `--storage-recover-on-startup`,
  `--auth-module-create-missing-user`, `--auth-module-create-missing-role`,
  `--auth-module-manage-roles`, `--storage-parallel-index-recovery`.
[#1916](https://github.com/memgraph/memgraph/pull/1916)

#### New features and improvements

- Implemented lazy evaluation of replication DNS. When using DNS in replication,
its resolution is postponed until streaming data using RPC connections,
improving efficiency in replication setups. The replication durability version
has been bumped to V4, and the corresponding migration function has been added.
[#2060](https://github.com/memgraph/memgraph/pull/2060)

- [`SHOW INSTANCES`](https://memgraph.com/docs/clustering/high-availability#show-instances) now returns
non-resolved IP addresses when using DNS in an HA cluster. DNS is resolved every
time data is streamed using RPC connections, ensuring seamless reconnections in
Kubernetes clusters even if a Pod is restarted. The `IPFamily` attribute is no
longer persisted as part of the `Endpoint` class.
[#2051](https://github.com/memgraph/memgraph/pull/2051)

- A new flag,
  [`--nuraft-log-file`](https://memgraph.com/docs/database-management/configuration#high-availability),
has been added to explicitly manage logs coming from NuRaft.
[#2084](https://github.com/memgraph/memgraph/pull/2084)

- Coordinator won't start if the `--instance-down-timeout-sec` option is smaller
than the `--instance-health-check-frequency-sec` option, ensuring that
[configuration options](https://memgraph.com/docs/database-management/configuration#high-availability)
for health pings are valid before starting the Memgraph coordinator instance.
[#2100](https://github.com/memgraph/memgraph/pull/2100)

- Validation added for the `experimental-enabled` flag. Users can now view logs
detailing which experimental features are enabled.
[#2106](https://github.com/memgraph/memgraph/pull/2106)

- In addition to specifying log file paths via the `--nuraft-log-file flag`,
users can now set the `MEMGRAPH_NURAFT_LOG_FILE` [environment
variable](https://memgraph.com/docs/database-management/configuration#environment-variables) to [control
the location of NuRaft logs](https://memgraph.com/docs/clustering/high-availability).
[#2113](https://github.com/memgraph/memgraph/pull/2113)

- Users can now see who is the leader and who are the followers in the output of
[`SHOW INSTANCES`](https://memgraph.com/docs/clustering/high-availability#show-instances). This
enhancement provides clearer visibility into the roles of instances within a
high-availability cluster.
[#2126](https://github.com/memgraph/memgraph/pull/2126)

- Memgraph now [supports Single
Sign-On](https://memgraph.com/docs/database-management/authentication-and-authorization/auth-system-integrations)
(SSO) modules. This feature allows for custom authentication schemes, enabling
users to log in using user-defined authentication modules. Upon receiving a Bolt
message for authorization, Memgraph will utilize the specified auth module for
the login process. The `--auth-module-executable` flag has been deprecated and
replaced by `--auth-module-mappings=basic`.
[#1990](https://github.com/memgraph/memgraph/pull/1990)

- Users can now connect to Memgraph [without specifying a
database](https://memgraph.com/docs/database-management/multi-tenancy). If a user has a default database,
they will connect to that database. If not, they will connect without a database
defined, allowing only database-less queries. Users can also define the database
via the client, executing each query against the specified database.
Multi-tenant authentication will be executed upon the first query execution.
[#1884](https://github.com/memgraph/memgraph/pull/1884)

- Improved query optimization by unifying multiple property-based filters into a
single filter that applies both upper and lower bounds.
[#2055](https://github.com/memgraph/memgraph/pull/2055)

- Introduced the USING HOPS LIMIT x syntax, enabling users to limit the number
of hops (edges crossed) during [graph
traversals](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal). This enhancement helps in
measuring and controlling the amount of work done in the traversal. This feature
covers traversals in the case of BFS, DFS and simple expand.
[#1928](https://github.com/memgraph/memgraph/pull/1928)

- Peak memory usage during the entire Memgraph run is now recorded and available
in [`SHOW STORAGE
INFO`](https://memgraph.com/docs/fundamentals/storage-memory-usage#inspect-memory-usage) and the metrics
server. This allows users to monitor memory usage more effectively and helps in
optimizing performance.
[#2048](https://github.com/memgraph/memgraph/pull/2048)

- Memgraph now supports an external module that enables Single Sign-On (SSO)
using the [SAML
protocol](https://memgraph.com/docs/database-management/authentication-and-authorization/auth-system-integrations#saml).
This module validates SAML responses and supports identity providers such as
Microsoft Entra ID and Okta.
[#2049](https://github.com/memgraph/memgraph/pull/2049)

- Added support for Single Sign-On (SSO) using [OIDC and OAuth2.0
protocol](https://memgraph.com/docs/database-management/authentication-and-authorization/auth-system-integrations#openid-connect-oidc).
[#2045](https://github.com/memgraph/memgraph/pull/2045)

- The support for the [Enum datatype](https://memgraph.com/docs/fundamentals/data-types#enum) has been
added. Enums can be used to enhance query performance and reduce memory usage
compared to string representations.
[#2039](https://github.com/memgraph/memgraph/pull/2039)

- Add support for [edge type property index](https://memgraph.com/docs/fundamentals/indexes).
[#1964](https://github.com/memgraph/memgraph/pull/1964)

- Added functionality that allows each user to [change their own
password](https://memgraph.com/docs/database-management/authentication-and-authorization/users).
[#2097](https://github.com/memgraph/memgraph/pull/2097)

- Removed support for Amazon Linux 2, CentOS 7 and Red Hat 7  [Linux
distributions](https://memgraph.com/docs/getting-started/install-memgraph/direct-download-links).
[#2165](https://github.com/memgraph/memgraph/pull/2165)

#### Bug fixes

- [Metrics can now be fetched](https://memgraph.com/docs/database-management/monitoring) via the HTTP
server during runtime if a Memgraph Enterprise license is enabled.
[#2067](https://github.com/memgraph/memgraph/pull/2067)

- Removed redundant duplicate label filtering.
[#2114](https://github.com/memgraph/memgraph/pull/2114)

- Fixed an issue where filters following a Cartesian operation were not always
correctly embedded into index scans, causing some query plans to miss a
Cartesian-to-Join swap.
[#1924](https://github.com/memgraph/memgraph/pull/1924)

- Resolved an issue where constants were incorrectly used as grouping keys in
aggregation queries.
[#2095](https://github.com/memgraph/memgraph/pull/2095)

- Resolved an issue where the routing table request logic inaccurately
identified the MAIN instance when the previous MAIN was not operational.
[#2133](https://github.com/memgraph/memgraph/pull/2133)

- Fixed an issue where the HA cluster would deadlock if an instance was down
when registered and it was the only instance in the cluster.
[#2135](https://github.com/memgraph/memgraph/pull/2135)

- This fix addresses a replication bug that previously prevented data and
metadata deltas from being in the same transaction. Metadata deltas are now sent
first to ensure the replica uses unique access, making automatic index creation
fully compatible with high availability and replication capabilities.
[#2125](https://github.com/memgraph/memgraph/pull/2125)

### MAGE v1.18 - Jul 3, 2024

#### New features and improvements

- Added a PostgreSQL [migration module to
MAGE](https://memgraph.com/docs/advanced-algorithms/available-algorithms/migrate) that enables direct
migration from a PostgreSQL database to Memgraph. This new module streamlines
the data import process by reducing the need for intermediate steps, such as
using CSV files. Users can now migrate tables, including nodes and
relationships, directly into Memgraph.
[#464](https://github.com/memgraph/mage/pull/464)

#### Bug fixes

- Fixed building of newly created Rust query modules. Previously, there was a
logical error in the code that prevented new Rust query modules from being built
correctly.
[#483](https://github.com/memgraph/mage/pull/483)

### Lab v2.15.0 - Jul 3, 2024

#### New features and improvements

- Share a query with other users with access to the same
  Lab and Memgraph. Works only on Lab in Docker and with
  enterprise license.
- You can configure quick-connect to connect to Memgraph
  with an encrypted connection with the new environment
  variable `QUICK_CONNECT_MG_IS_ENCRYPTED`.
- The size of the Lab Docker image has been reduced to
  one-fifth of its original size.

#### Bug fixes

- Boolean `false` value is now shown correctly in the expanded
  row of query table results.
- Passwords are now hidden in the run history when running a
  query to create a new user with a password or change the
  user's password.

### Lab v2.14.0 - May 23, 2024

#### New features and improvements

- Add SSO (single sign-on) configuration for Microsoft Entra ID and/or Okta with
  SAML and/or OAuth standard protocols (client only).
- Run a query from a query collection in any of the vertical
  and/or horizontal views with a selection dropdown next to
  the "Run query" button.
- For each query, query performance details will be in the
  query summary on how much time was spent on each part of
  your query: plan, parsing, execution and full roundtrip.
- `ZonedDateTime` data type will be presented with timezone
  information in the query data results table.

#### Bug fixes

- A bug where you would be locked in the fullscreen mode after
  CSV import usage has been fixed.

### Memgraph v2.17.0 - May 22, 2024

#### Breaking changes

- The flag
[`--replication-restore-state-on-startup`](https://memgraph.com/docs/database-management/configuration)
is now set to `true` by default.
[#2031](https://github.com/memgraph/memgraph/pull/2031)

- Memgraph will now fail to start if the Bolt server cannot be created, ensuring
a dedicated port is available. Failures in creating the WebSocket and HTTP
servers will no longer crash Memgraph but will be logged more effectively. If
the WebSocket server fails, the `LoggerSink` will not be added to it.
Additionally, Memgraph will crash if the authentication folder cannot be
created.
[#2017](https://github.com/memgraph/memgraph/pull/2017)

#### New features and improvements

- Added functionality to [create a
user](https://memgraph.com/docs/database-management/authentication-and-authorization/users) or
[role](https://memgraph.com/docs/database-management/authentication-and-authorization/role-based-access-control)
only if it does not already exist. This update prevents errors when creating a
user or role that already exists.
[#2032](https://github.com/memgraph/memgraph/pull/2032)

- Introduced Docker images specifically tailored for
[debugging](https://memgraph.com/docs/database-management/debugging), with Memgraph built in the
`RelWithDebInfo mode`. These containers come equipped with essential debugging
tools such as `perf`, `gdb`, and `pgrep`, along with other useful apt packages.
[#2047](https://github.com/memgraph/memgraph/pull/2047)

- Added the capability to count the number of hops in queries for each
expansion. Each unit of work, including filtering on edge types, edge
properties, or visiting already processed nodes, is counted as a hop. The
total number of hops is returned in the `number_of_hops` field in the query
summary.
[#1935](https://github.com/memgraph/memgraph/pull/1935)

- Introduced a new small vector datatype (`16B`) with a small buffer
optimization. This new datatype is used for `Vertex` to have [smaller labels
collection and smaller edge collections](https://memgraph.com/docs/fundamentals/storage-memory-usage),
resulting in a `24B` reduction per `Vertex`.
[#1970](https://github.com/memgraph/memgraph/pull/1970)

#### Bug fixes

- Data instances now use `0.0.0.0:replication_port` for creating the replication
server instead of the instance's IP address. This update ensures that the
replication server starts correctly.
[#2038](https://github.com/memgraph/memgraph/pull/2038)

- The replication stream is now handled as a local variable for transactions,
preventing assertion failures during recovery.
[#2018](https://github.com/memgraph/memgraph/pull/2018)

- Implemented changes to discard leadership requests on follower instances. This
update ensures that followers no longer accept user actions intended for the
leader.
[#1932](https://github.com/memgraph/memgraph/pull/1932)

- Fixed a NuRaft launcher race condition. The raft server is now considered
successfully created once it behaves as a leader.
[#2040](https://github.com/memgraph/memgraph/pull/2040)

### Memgraph v2.16.1 - May 15, 2024

#### Breaking changes

- Replicate only durable commits to replicas during recovery. This ensures that
  during the recovery process, replicas only receive commits that have been
  fsynced on the main node, preventing the replication of non-durable data and
  improving overall system consistency.
  [#1991](https://github.com/memgraph/memgraph/pull/1991)

#### New features and improvements

- Added exponential backoff to the coordinator force reset function. This
  improvement helps reduce pressure on logs and allows time for machines to come
  back online by implementing a backoff strategy that starts at 1 second and
  increases to 5 seconds, followed by resets every 5 seconds.
  [#1996](https://github.com/memgraph/memgraph/pull/1996)

- Upgraded the Clojure client dependencies. Memgraph now uses
  `gorillabs/neo4j-clj` version 5.0.0 for Jepsen testing, Clojure version
  1.11.2, and `neo4j-harness` version 5.10.0.
  [#1987](https://github.com/memgraph/memgraph/pull/1987)

- Added timezone support to Memgraph with the new [zoned datetime
  type](https://memgraph.com/docs/fundamentals/data-types#zoneddatetime). This feature allows you to
  represent points in time with a defined timezone, enabling you to work with
  temporal data from different locales and easily migrate datasets with zoned
  values. [#1866](https://github.com/memgraph/memgraph/pull/1866)
  [#1976](https://github.com/memgraph/memgraph/pull/1976)

- Added `last_succ_resp_ms` and [coordinator health
  status](https://memgraph.com/docs/configuration/high-availability) to the output of the `SHOW
  INSTANCES` command. This update allows users to see the elapsed time in
  milliseconds since the last successful response from instances to the leader's
  health ping, as well as the health status of the coordinator.
  [#1985](https://github.com/memgraph/memgraph/pull/1985)

- Added environment variables for HA Kubernetes cluster setup. This update
  allows starting the coordinator and data instances using environment
  variables, enhancing flexibility and ease of configuration for [high
  availability](https://memgraph.com/docs/configuration/high-availability) clusters in Kubernetes
  environments. [#1951](https://github.com/memgraph/memgraph/pull/1951)

- Added an option to register a replica after the main instance has already been
  registered in a [high availability](https://memgraph.com/docs/configuration/high-availability) (HA)
  setup. This feature enhances the flexibility of replica management in HA
  configurations. [#1911](https://github.com/memgraph/memgraph/pull/1911)

- Improved the output of the `SHOW INSTANCES` command. Changes include renaming
  `raft_socket_addr` to `coord_server` and `coordinator_socket_addr` to
  `mgmt_server`, and [adding the Bolt server to the
  output](https://memgraph.com/docs/configuration/high-availability).
  [#1965](https://github.com/memgraph/memgraph/pull/1965)

#### Bug fixes

- Resolved an issue in the automatic index creation process where transactions
  creating multiple entities simultaneously could lead to unexpected behaviors.
  This patch ensures reliable auto index creation when dealing with multiple
  entity creations in a single transaction.
  [#1926](https://github.com/memgraph/memgraph/pull/1926)

- This fix ensures robust index rewriting and prevents potential core dumps in
  situations involving pattern comprehension within queries. Previously, the
  application could fail when using label or property-label indices with pattern
  comprehensions.
  [#1968](https://github.com/memgraph/memgraph/pull/1968)

- Addressed a flaw where skipped timestamps were not finalized on replicas,
  leading to inaccuracies in the oldest active timestamp information provided to
  the GC. This fix ensures all skipped timestamps are now marked as finalized,
  allowing the GC to accurately manage memory and clear Deltas as needed.
  Additionally, the RPC session buffer now resets to a maximum of 4 MiB
  post-session execution, improving memory management.
  [#2003](https://github.com/memgraph/memgraph/pull/2003)

- Fixed the DNS resolution capability for endpoint parsing. This update resolves
  issues with DNS resolution when registering replicas and/or coordinator
  instances. [#2004](https://github.com/memgraph/memgraph/pull/2004)

- Expose coordinator's internal functionality to manage the cluster. Users can
  now execute actions such as `DEMOTE INSTANCE` and `FORCE RESET CLUSTER STATE`,
  which were previously internal operations. These functionalities allow users
  to reset cluster state and promote their chosen instance to be the MAIN.
  [#1981](https://github.com/memgraph/memgraph/pull/1981)

### Memgraph v2.16.0 - Apr 10, 2024

#### New features and improvements

- Added a new command, [`DROP GRAPH`](https://memgraph.com/docs/querying/clauses/drop-graph), that
  deletes data in the database in an efficient manner.
  [#1723](https://github.com/memgraph/memgraph/pull/1723)
- Fixed a bug that caused the incorrect coordinator to become the leader due to
  missing cluster state data.
  [#1897](https://github.com/memgraph/memgraph/pull/1897)
- Added support for [edge type indices](https://memgraph.com/docs/fundamentals/indexes#edge-type-index)
  to optimize query performance on graphs.
  [#1542](https://github.com/memgraph/memgraph/pull/1542)
- Added [automatic index creation](https://memgraph.com/docs/fundamentals/indexes) for node labels and
  edge types. Two new configuration flags control this feature for labels and
  edge-types automatically encountered by the database.
  [#1841](https://github.com/memgraph/memgraph/pull/1841)
- Added support for `bolt+routing` when connecting to a cluster of Memgraph
  instances. [#1796](https://github.com/memgraph/memgraph/pull/1796)
- Expanded the [`min()` and `max()`](https://memgraph.com/docs/querying/functions#aggregation-functions)
  functions to support temporal types, improving date and time manipulation
  capabilities.
  [#1790](https://github.com/memgraph/memgraph/pull/1790)
- Renamed [HA flags](https://memgraph.com/docs/configuration/configuration-settings#high-availability)
  `--raft-server-id` to `--coordinator-id`, `--raft-server-port` to
  `--coordinator-port`, `--coordinator-server-port` to `--management-port`
  to improve cluster setup.
  [#1820](https://github.com/memgraph/memgraph/pull/1820)
  - Reduced the size of the Delta from 80B to 56B
  [#1747](https://github.com/memgraph/memgraph/pull/1747)
- Extended the C/C++ API to support [query
  execution](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api#executionheaders).
  [#1773](https://github.com/memgraph/memgraph/pull/1773)
- [Updated queries](https://memgraph.com/docs/configuration/high-availability) used for registering
  replication instances and coordinator instances.
  [#1809](https://github.com/memgraph/memgraph/pull/1809)
- [Introduced the lock on operations](https://memgraph.com/docs/configuration/high-availability) such as
  `SETTING INSTANCE` to MAIN, `REGISTER INSTANCE`, `UNREGISTER INSTANCE`, and
  `FAILOVER` for better control of the cluster state.
  [#1819](https://github.com/memgraph/memgraph/pull/1819)
- When doing failover, the Coordinator [chooses a new
  MAIN](https://memgraph.com/docs/configuration/high-availability) by comparing which `memgraph` database
  is up-to-date. [#1729](https://github.com/memgraph/memgraph/pull/1729)
- [Changing instance's replication role](https://memgraph.com/docs/configuration/high-availability) is
  now backed by Raft. [#1728](https://github.com/memgraph/memgraph/pull/1728)
- If REPLICA diverges from MAIN, MAIN will automatically initiate a forced data
  reset on REPLICA. [#1777](https://github.com/memgraph/memgraph/pull/1777)
- User can create, set, or remove labels using property values.
  [#1762](https://github.com/memgraph/memgraph/pull/1762)
- Introduced the ability to [fetch edges based on edge
  ID](https://memgraph.com/docs/configuration/configuration-settings#storage).
  [#1808](https://github.com/memgraph/memgraph/pull/1808)
- Added support for [pattern
  comprehensions](https://memgraph.com/docs/querying/expressions#pattern-comprehension) when used in
  `WITH` and  `RETURN` clauses.
  [#1827](https://github.com/memgraph/memgraph/pull/1827)
  [#1874](https://github.com/memgraph/memgraph/pull/1874)
  [#1903](https://github.com/memgraph/memgraph/pull/1903)
- Introduced a [force reset](https://memgraph.com/docs/configuration/high-availability) feature to revert
  the cluster to its pre-operation state in case of failure.
  [#1836](https://github.com/memgraph/memgraph/pull/1836)

#### Bug fixes

- Added an early exit when doing `OR` filter expressions if the first expression
  evaluates to `true`.
  [#1738](https://github.com/memgraph/memgraph/pull/1738)
- Configuration settings are now properly verified to ensure that periodic
  snapshots are only taken when appropriate.
  [#1835](https://github.com/memgraph/memgraph/pull/1835)
- Improved replication stability in high-availability (HA) clusters to handle
  situations where the REPLICA node is several epochs or commits behind the MAIN
  node.
  [#1743](https://github.com/memgraph/memgraph/pull/1743)
- Fixed a bug that could occur when shutting down the Raft server. Initializing
  Memgraph enterprise instance without specifying any flag used in
  high-availability is now allowed.
  [#1823](https://github.com/memgraph/memgraph/pull/1823)
- Query allocations now use a single allocation subsystem rather than two.
  Memory no longer horded by bolt workers.
  [#1801](https://github.com/memgraph/memgraph/pull/1801)
- The [`schema.node_type_properties()` and `schema.rel_type_properties()`](https://memgraph.com/docs/querying/schema)
  functions now work as equivalents to Neo4j's functions.
  [#1718](https://github.com/memgraph/memgraph/pull/1718)
- Merging node or edge with [null property now raises an
  error](https://memgraph.com/docs/querying/clauses/merge#41-merging-a-null-property).
  [#1810](https://github.com/memgraph/memgraph/pull/1810)
- When adding coordinators in cluster, action will fully succeed or fail
  before coordinators returns result from query.
  [#1792](https://github.com/memgraph/memgraph/pull/1792)
- Fixed a bug responsible for the incorrect edge count when a graph was dropped
  using `DROP GRAPH`. Now, accurate edge counts are displayed in the `SHOW
  STORAGE INFO` output. [#1921](https://github.com/memgraph/memgraph/pull/1921)

### MAGE v1.16 - Apr 10, 2024

#### New features and improvements

- Added [md5
  function](https://memgraph.com/docs/advanced-algorithms/available-algorithms/util_module#md5) along
  with the procedure.
  [#444](https://github.com/memgraph/mage/pull/444)
- Added [`set_property`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/set_property).
  [#445](https://github.com/memgraph/mage/pull/445)
- Added query module for [periodic
  delete](https://memgraph.com/docs/advanced-algorithms/available-algorithms/periodic#delete) of graph
  entities.
  [#446](https://github.com/memgraph/mage/pull/446)
- Added the [`date.add` query
  function](https://memgraph.com/docs/advanced-algorithms/available-algorithms/date#add) for more
  convenience in handling temporal values.
  [#456](https://github.com/memgraph/mage/pull/456)

#### Bug fixes

- `Periodic` and `do` modules now work with the Memgraph API without the need
  for a user to authenticate [#461](https://github.com/memgraph/mage/pull/461)

### Lab v2.13.0 - Apr 10, 2024

#### New features and improvements

- Select a color for your connections, enhancing distinction within
  the Lab web interface.
- Use precise CSV imports by defining custom line filters,
  and processing specific file lines based on user-defined conditions.
- Filter datasets by category such as "Knowledge Graphs", "Supply Chain",
  and others.
- Run the Cypher query from the query parameters tab without switching
  back to the Cypher query tab in the query editor.
- Enhanced CSV import now provides extensive notices, best practices
  and warnings tailored to the CSV import process and specific data
  types such as dates and date times, aiding you in smoother data handling.

#### Bug fixes

- Queries in a collection from a European backpacking dataset now return
  a correct response.
- Running a query in a horizontal view now works as expected.
- A bug introduced in v2.12.0 with failed key encryption when
  adding OpenAI and Azure OpenAI keys is fixed now.

### Memgraph v2.15.2 - Apr 6, 2024

#### Bug fixes

- Fixed wrong [text search](https://memgraph.com/docs/configuration/text-search) module name.
  [#1900](https://github.com/memgraph/memgraph/pull/1900).

### Memgraph v2.15.1 - Mar 29, 2024

#### New features and improvements

- Introduced a new flag,
  [`storage-delta-on-identical-property-update`](https://memgraph.com/docs/configuration/configuration-settings#storage),
  which allows for the optional creation of a delta object when a property is
  updated with the same value.
  [#1791](https://github.com/memgraph/memgraph/pull/1791).

- Added durable text indices and [basic text search](https://memgraph.com/docs/configuration/text-search)
capabilities. [#1603](https://github.com/memgraph/memgraph/pull/1603)
[#1739](https://github.com/memgraph/memgraph/pull/1739)

- Removed support for Debian 10, Fedora 36 and Ubutnu 18.04. Added
support for Debian 12, Debian 12 ARM, Fedora 38, Fedora 39 and Rocky 9.3 [Linux
distributions](https://memgraph.com/docs/getting-started/install-memgraph/direct-download-links).
[#1814](https://github.com/memgraph/memgraph/pull/1814)

#### Bug fixes

- Resolved an issue where snapshot and GC operations would block the initiation
of new transactions. [#1759](https://github.com/memgraph/memgraph/pull/1759)

- Reading/writing snapshots on coordinators now works as expected.
[#1788](https://github.com/memgraph/memgraph/pull/1788)

### Lab v2.12.2 - Mar 8, 2024

#### Bug fixes

- Expand, collapse and hide actions on nodes are working as expected,
  as opposed to the bug experienced in v2.12.x.

### Lab v2.12.1 - Mar 1, 2024

#### Bug fixes

- An error message for empty and invalid GSS is only shown on
  results graph view.

### Lab v2.12.0 - Feb 28, 2024

#### New features and improvements

- Change the color of nodes and edges in the graph view using
  the color picker.
- Change the nodes and edges label shown in the graph view to
  any of the following: labels, property values or none.
- Along with OpenAI, you can now setup remote Azure OpenAI or local
  Ollama LLM models for your GraphChat.
- Fine-tune permissions of the GraphChat to run only read, create,
  update and/or delete Cypher commands.
- A quick link to GraphChat settings is now in the upper right corner
  of GraphChat conversation.
- The query request limit is increased from 100kB to 20MB to enable you to
  run a lot more queries at once from the query execution editor.

#### Bug fixes

- Query execution editor loads the previous state when opened again, as
  opposed to the empty state bug experienced in v2.11.1.
- You can now select a property type for source and target nodes when
  creating edges via CSV import.
- Unselected CSV column in CSV import will not result in CSV error
  anymore.
- Current database in a multi-tenancy environment is shown correctly
  for Memgraph version >= 2.15.

### Memgraph v2.15.0 - Feb 28, 2024

#### Upgrading to Memgraph v2.15.0

> **Info**
>
> Starting from Memgraph version 2.15.0, password hashing
> has been fortified with salt for improved security. Previously, passwords were
> hashed using SHA256 without salt. Please note that bcrypt, which is the default
> hashing algorithm, has always incorporated salt for enhanced security.
>
> - **Automatic Migration:** Passwords hashed without salt will be automatically
>   reverted to the saltless algorithm upon upgrade. Subsequent successful logins
>   trigger a rehashing process with salt, updating the data on disk.
> - **Seamless Transition:** Most users will experience a seamless transition
>   without any manual intervention. A user's first login post-upgrade triggers
>   the hash update.
> - **Migration Considerations:** Upon starting Memgraph version 2.15.0 for the
> first time, ensure that the `--password-encryption-algorithm` flag is correctly
> set to match the actual hashing algorithm used. This is crucial for maintaining
> authentication data integrity.

#### New features and improvements

- Enhanced security by integrating a password hashing algorithm into
  authentication durability mechanisms, alongside existing password hashes.
  Passwords hashed with SHA-256 will be rehashed with salt upon validation for
  improved security.
  [#1644](https://github.com/memgraph/memgraph/pull/1644)

- Introduced a function `propertySize` to calculate the size of properties (in
  bytes) within Memgraph storage. For more details, refer to the
  [Functions documentation](https://memgraph.com/docs/querying/functions).
  [#1557](https://github.com/memgraph/memgraph/pull/1557)

- The process of demoting old MAIN to REPLICA should eventually succeed fully
  without any issues.
  [#1711](https://github.com/memgraph/memgraph/pull/1711)

- The coordinator now aligns REPLICA to consistently listen to the correct MAIN.
  [#1711](https://github.com/memgraph/memgraph/pull/1711)

- Improved memory handling of Deltas. The value for Deltas is changed from  104B
  to 80B. For more details, refer to the [Storage memory usage
  documentation](https://memgraph.com/docs/fundamentals/storage-memory-usage).
  [#1688](https://github.com/memgraph/memgraph/pull/1688)

- Added support for query parameters in LOAD CSV. This enhancement allows users
  to dynamically specify file paths in the LOAD CSV clause. Executing queries
  like `LOAD CSV FROM $file AS row CREATE (:Node { column: row[0] });` is now
  possible, where `$file` is a variable representing a valid path to a CSV file.
  For more details, refer to the [Expressions documentation](https://memgraph.com/docs/querying/expressions).
  [#1653](https://github.com/memgraph/memgraph/pull/1653)

- Added support for query parameters in RETURN LIMIT. Users can now specify the
  limit in a RETURN statement through a parameter, such as `MATCH (n) RETURN n
  LIMIT $limit;`. This addition makes it easier to dynamically adjust the number
  of results returned by a query. For more details, refer to the [Expressions
  documentation](https://memgraph.com/docs/querying/expressions).
  [#1654](https://github.com/memgraph/memgraph/pull/1654)

- [Enterprise Ed.] Introduced an experimental feature flag for system
  replication (`--experimental-enabled=system-replication`). This new flag enables both
  authentication and multi-tenancy replication.

- Introduced enhancements for multi-tenancy support. The new `SHOW DATABASE;`
  command reveals the currently used database, returning `NULL` if none is
  active. `SHOW DATABASES;` has been updated to display only the existing set of
  multitenant databases. For more details, refer to the [Multi-tenancy
  documentation](https://memgraph.com/docs/configuration/multi-tenancy).
  [#1550](https://github.com/memgraph/memgraph/pull/1550)

- When using replication, flag `--replication-restore-state-on-startup` needs to
  be set to `true`.  For more details, refer to the [Configuration settings
  documentation](https://memgraph.com/docs/configuration/configuration-settings).
  [#1707](https://github.com/memgraph/memgraph/pull/1707)

- REPLICAs are now configured to listen exclusively to a single MAIN instance.
  In high availability (HA) setups, the coordinator now manages the assignment
  of REPLICA instances. For configurations using replication without HA, the
  MAIN instance directly communicates its ID to each REPLICA. For more details,
  refer to the [High availability
  documentation](https://memgraph.com/docs/configuration/high-availability).
  [#1674](https://github.com/memgraph/memgraph/pull/1674)

- Added coordinator instance for managing HA cluster. For more details, refer to
  the [High availability documentation](https://memgraph.com/docs/configuration/high-availability).
  [#1608](https://github.com/memgraph/memgraph/pull/1608)

- Added support for restart of replication instances. MAIN can restart as MAIN
  and as REPLICA depending on the cluster state in the moment of restart.
  REPLICA always stays REPLICA. For more details, refer to the [High
  availability documentation](https://memgraph.com/docs/configuration/high-availability).
  [#1672](https://github.com/memgraph/memgraph/pull/1672)

- Added automatic failover from a single coordinator. For more details refer to
  [High availability documentation](https://memgraph.com/docs/configuration/high-availability).
  [#1646](https://github.com/memgraph/memgraph/pull/1646)

- Added flags `--raft-server-id` and `--raft-server-port` for creating
  coordinators. For more details, refer to the [High availability
  documentation](https://memgraph.com/docs/configuration/high-availability).
  [#1687](https://github.com/memgraph/memgraph/pull/1687)

- Improved the authentication module with better user and role handling mechanisms. For more
  details, refer to the [Auth module documentation](https://memgraph.com/docs/configuration/auth-module).
  Key enhancements include:
  - Elimination of automatic user creation by the auth module
    - Introduction of checks to determine the presence of users or the use of an
      auth module before proceeding
  - Support for roles with database access
  - The `authenticate()` function now returns a user or role
  - When an auth module is in use, no data is locally saved
  - AuthChecker now generates `QueryUser` (user or role), which is responsible
    for actual authorization checks
  - Deprecated the following configuration flags:
    `--auth-module-create-missing-user`,  `--auth-module-create-missing-role`,
    `--auth-module-manage-roles`
    [#1699](https://github.com/memgraph/memgraph/pull/1699)
  - Introduced multi-tenancy and system information to [SHOW
  REPLICAS](https://memgraph.com/docs/configuration/replication).
  [#1735](https://github.com/memgraph/memgraph/pull/1735)

#### Bug fixes

- Fixed an issue with the memory tracker not accurately counting memory usage
  after exceeding the memory limit.
  [#1651](https://github.com/memgraph/memgraph/pull/1651)

- Resolved an issue that led to a crash when an unbound variable was used inside a subquery.
  [#1710](https://github.com/memgraph/memgraph/pull/1710)

- Fixed deadlocks in jemalloc caused by the memory tracker.
  [#1715](https://github.com/memgraph/memgraph/pull/1715)

- Fixed an issue in the Breadth-First Search (BFS) expansion process where the
  path was not correctly restored to its previous state after expansion, unless
  it was modified.  The path is now appropriately shrunk and the state is
  restored to its prior condition only when changes have occurred.
  [#1745](https://github.com/memgraph/memgraph/pull/1745)

### MAGE v1.15.0 - Feb 28, 2024

#### Bug fixes

- Community detection now respects memory limits and doesn't the crash database.
  [#431](https://github.com/memgraph/mage/pull/431)

### Lab v2.11.1 - Jan 22, 2024

#### New features and improvements

- The flow of the query plan is now easily tracked thanks to the arrows that connect individual steps.

#### Bug fixes

- Importing data from a new CYPHERL file format that contains queries for
  creating triggers now works as expected.
- Importing a large CYPHERL file with more than 1M lines doesn't
  crash the Lab anymore.

### Memgraph v2.14.1 - Feb 16, 2024

#### Bug fixes

- Fixed an issue with accumulated path evaluation affecting built-in algorithms:
  - Depth-first search (DFS)
  - Breadth-first search (BFS)
  - Weighted shortest path (WSP)
  - All shortest paths (ASP).
  [#1642](https://github.com/memgraph/memgraph/pull/1642)

### Memgraph v2.14.0 - Jan 22, 2024

#### New features and improvements

- [`toBoolean()` function](https://memgraph.com/docs/querying/functions#scalar-functions) now maps
  strings `"true"` and `"t"` to `true` and `"false"` and `"f"` to `false`.
  [#1620](https://github.com/memgraph/memgraph/pull/1620)
- Labels can now also be defined using query
  [parameters](https://memgraph.com/docs/querying/expressions#parameters).
  [#1602](https://github.com/memgraph/memgraph/pull/1602)
- [`toString()` function](https://memgraph.com/docs/querying/functions#scalar-functions) returns double
  values with precision 15. [#1576](https://github.com/memgraph/memgraph/pull/1576)
- When [dumping a
  database](https://memgraph.com/docs/configuration/data-durability-and-backup#database-dump), triggers
  are also included in the CYPHERL file.
  [#1610](https://github.com/memgraph/memgraph/pull/1610)

#### Bug fixes

- The [`--storage-mode` configuration
  flag](https://memgraph.com/docs/configuration/configuration-settings#storage) now works as expected.
  [#1609](https://github.com/memgraph/memgraph/pull/1609)
- Memgraph returns an error instead of crashing when trying to execute the
  [`SHOW STORAGE INFO;`](https://memgraph.com/docs/fundamentals/storage-memory-usage#check-storage-mode)
  query without the required permission.
  [#1566](https://github.com/memgraph/memgraph/pull/1566)
- Memgraph returns an error instead of crashing when aborting a transaction due
  to lack of memory. [#1589](https://github.com/memgraph/memgraph/pull/1589)
- [REPLICA](https://memgraph.com/docs/configuration/replication) timestamp is now correctly propagated.
  [#1615](https://github.com/memgraph/memgraph/pull/1615)
- Index hinting now works as expected. [#1606](https://github.com/memgraph/memgraph/pull/1606)
- Using the path identifier from the `CREATE` clause in other parts of the query
  will no longer crash the database. For example, the following query will
  execute successfully:

```
CREATE p0=()-[:T0]->()
MERGE ({k:(size(p0))})
RETURN 1;
```
  [#1629](https://github.com/memgraph/memgraph/pull/1629)

### MAGE v1.14.0 - Jan 22, 2024

- Internal improvements.

### Lab v2.11.0 - Jan 3, 2024

#### New features and improvements

- Now, you can import CSV files using the CSV import configuration tool instead
  of writing Cypher queries.
- The query plan chart is now available in the *Summary* tab for each query run.
- You can profile a query plan to enrich the query plan chart with the absolute
  and relative execution time spent on each query plan step.
- The generated query plan can be downloaded in the JSON format.
- LLM GraphChat now works even when connected to Neo4j.
- Query data view will be shown by default if there are non-graph data in the
  query results.

#### Bug fixes

- Hiding nodes in the *Graph results* view now works as expected.
- Lists in the notifications are now shown with correct
  indentation.

### MAGE v.1.13.0 - Dec 8 2023

- The [`regexGroups`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/text#regexgroups)
  procedure of the `text` query module enables the search for regex
  subexpression in a text using the [C++
  regex](https://en.cppreference.com/w/cpp/regex) library. and the
  [`format()`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/text#format) procedure
  enables formatting string using the [fmt
  library](https://github.com/fmtlib/fmt).
  [#420](https://github.com/memgraph/mage/pull/420)
  [#423](https://github.com/memgraph/mage/pull/423)
- Deleted return values for query procedures and functions ran in [the in-memory
  analytical storage mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#implications) are
  now properly handled. [#400](https://github.com/memgraph/mage/pull/400)

### Memgraph v2.13.0 - Dec 8, 2023

#### Upgrading to Memgraph v2.13.0

> **Info**
>
> Existence and unique [constraints](https://memgraph.com/docs/fundamentals/constraints) can now also be
> recovered in parallel, just like [indexes](https://memgraph.com/docs/fundamentals/indexes). That is why
> the `storage-parallel-index-recovery` configuration flag has been deprecated,
> and now you can enable this behavior using the
> [`storage-parallel-schema-recovery`
> configuration](https://memgraph.com/docs/configuration/configuration-settings#storage) flag.

#### New features and improvements

- Now you can check information about node labels and relationship types by
  running the [schema-related queries](https://memgraph.com/docs/querying/schema#schema-related-queries)
  `SHOW NODE_LABELS INFO;` and `SHOW EDGE_TYPES INFO;`.
  [#1466](https://github.com/memgraph/memgraph/pull/1466)
- When the [`schema.node_type_properties()` and the
  `schema.rel_type_properties()`
  procedures](https://memgraph.com/docs/querying/schema#schema-related-procedures) output the `mandatory`
  values, the `True` value indicates a given node or relationship type is
  associated with only one property, instead of indicating the existence of any
  properties on the node or relationship.
  [#1510](https://github.com/memgraph/memgraph/pull/1510)
- Existence and unique [constraints](https://memgraph.com/docs/fundamentals/constraints) can now also be
  recovered in parallel, just like [indexes](https://memgraph.com/docs/fundamentals/indexes). To enable
  this behaviour, set the [`storage-parallel-schema-recovery`
  configuration](https://memgraph.com/docs/configuration/configuration-settings#storage) flag to `true`.
  As the new configuration flag includes the constraints, the
  `storage-parallel-index-recovery` has been deprecated.
  [#1545](https://github.com/memgraph/memgraph/pull/1545)
- Now you can assert indexes and constraints using the
  [`schema.assert()`](https://memgraph.com/docs/querying/schema#schema-assert) procedure.
  [#1485](https://github.com/memgraph/memgraph/pull/1485)
- Improved the logic when checking [the `exist()`
  statements](https://memgraph.com/docs/querying/functions#pattern-functions) within [the `WHERE`
  clause](https://memgraph.com/docs/querying/clauses/where#17-filter-with-pattern-expressions) in order
  to improve performance.
  [#1539](https://github.com/memgraph/memgraph/pull/1539)
- The filter information message of `EXPLAIN` and `PROFILE` queries displays the
  exact filtering in the filter expression.
  [#1481](https://github.com/memgraph/memgraph/pull/1481)
- When the weight lambda of [Weighted Shortest Path and All Shortest Paths
  algorithms](https://memgraph.com/docs/advanced-algorithms/built-in-graph-algorithms)
  uses node properties, it will include all nodes of the path, even the start
  and end nodes. Also it's possible to mix node and relationships properties, by
  coalescing the relationship property in the weight lambda.
  [#1434](https://github.com/memgraph/memgraph/pull/1434)
- There have been changes to `mgp.hpp` method in the C++ API.
  [#1536](https://github.com/memgraph/memgraph/pull/1536)

#### Bug fixes

- If results of [aggregation
  functions](https://memgraph.com/docs/querying/functions#aggregation-functions)
  are returned grouped or if there is no input to aggregate over, rather than
  returing `0` or `empty` together with `NULL`, the functions won't return
  anything. [#1531](https://github.com/memgraph/memgraph/pull/1531)
- If the initialisation CYPHERL file passed using the
  [`--init-file`](https://memgraph.com/docs/configuration/configuration-settings#other)
  configuration flag tries to create a user that already exists, the exception
  will be caught and Memgraph will start successfully.
  [#1465](https://github.com/memgraph/memgraph/pull/1465)
- All deleted relationships are now correctly replicated to REPLICA instances.
  Also, replication failures no longer prevent MAIN to write into WAL files.
  [#1540](https://github.com/memgraph/memgraph/pull/1540)
- The cartesian evaluation of three MATCH clauses no longer uses sequential
  scanning, thus improving the performance of such queries.
  [#1555](https://github.com/memgraph/memgraph/pull/1555)
- A bug was fixed that would MATCH and RETURN wrong grouped aggregations.
  [#1518](https://github.com/memgraph/memgraph/pull/1518)
- Memgraph no longer throws a "Failed deallocation" error when working with the
  [`migrate` query module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/migrate).
  [#1492](https://github.com/memgraph/memgraph/pull/1492)
-  The [C, C++ and Python APIs](https://memgraph.com/docs/custom-query-modules) have been extended with
  functions to check what storage mode is being used and whether graph elements
  (or lists/maps thereof) have been deleted. They allow the proper handling of
  deleted return values or query procedures and functions ran in [the
  in-memory analytical storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#implications) are now properly
  handled.[#1395](https://github.com/memgraph/memgraph/pull/1395)
- A bug was fixed where Memgraph would send a success message to clients before
  certain database operations were executed successfully.
  [#1556](https://github.com/memgraph/memgraph/pull/1556)
- Correct exception type is now sent on failure of the BEGIN, COMMIT and
  ROLLBACK commands. [#1560](https://github.com/memgraph/memgraph/pull/1560)

### Lab v2.10.0 - Dec 8, 2023

#### New features and improvements

- OpenAI LLM integration enables you to query the database using the English language instead of writing Cypher queries.
- After each query run, query summary, statistics, and database notifications are now shown in the result summary
  tab.

### MAGE v1.12.1 - Nov 21, 2023

- Changes to internal utilities. [#415](https://github.com/memgraph/mage/pull/415)

### Memgraph v2.12.1 - Nov 17, 2023

#### Upgrading to Memgraph v2.12.1

> **Info**
>
> Cached query plans are no longer saved during the time interval defined by the
> `--query-plan-cache-ttl`, rather they are cached by count.
>
> This means that you need to replace any custom setting of the
> `--query-plan-cache-ttl` with the `--query-plan-cache-max-size` [configuration
> flag](https://memgraph.com/docs/configuration/configuration-settings#query). By default, the maximum
> number of cached query plans is
> 1000.
>
> This change has been made to reduce the memory usage of queries with `MERGE`
> queries and variables.

#### New features and improvements

- Cached query plans are no longer saved during the time interval defined by the
  `--query-plan-cache-ttl`, rather they are cached by count. You can set the
  maximum number of cached query plans using the `--query-plan-cache-max-size`
  [configuration flag](https://memgraph.com/docs/configuration/configuration-settings#query). By default,
  the maximum number of cached query plans is
  1000. This change has been made to reduce the memory usage of queries with
  `MERGE` queries and variables.
  [#1348](https://github.com/memgraph/memgraph/pull/1348)
- Procedures are no longer limited to 100 MB. They can use unlimited resources.
  If you want to [limit memory usage of
  procedures](https://memgraph.com/docs/custom-query-modules/manage-query-modules#controlling-procedure-memory-usage)
  add `PROCEDURE MEMORY LIMIT 100 KB` after calling a procedure and before
  yielding results. [#1506](https://github.com/memgraph/memgraph/pull/1506)
- The [`SHOW STORAGE INFO;`](https://memgraph.com/docs/configuration/server-stats) query now returns the
  non-swapped physical RAM memory a task has used (as reported by the operating
  system) as `memory_res`, not `memory_usage`. `memory allocated` has been
  replaced with `memory_tracked`, and it still defines the amount of RAM
  allocated in the system and tracked by Memgraph. The query also returns the
  number of memory-mapped areas that the kernel allows a process to have as
  `vm_max_map_count`. [#1408](https://github.com/memgraph/memgraph/pull/1408)
  [#1426](https://github.com/memgraph/memgraph/pull/1426)
- Now you can delete all nodes and relationships on a path by using the
  following `DETACH DELETE` syntax: [`MATCH p = (:X)-->()-->()-->() DETACH
  DELETE p;`](https://memgraph.com/docs/querying/read-and-modify-data#delete-data).
  [#1383](https://github.com/memgraph/memgraph/pull/1383)
- After
  [recovery](https://memgraph.com/docs/configuration/data-durability-and-backup#restore-data),
  the logs show which indexes and constraints have been recovered.
  [#1480](https://github.com/memgraph/memgraph/pull/1480)
- Information about using custom memory resources for handling memory when
  running has been removed from the [logs](https://memgraph.com/docs/configuration/logs) as it's
  unnecessary. [#1452](https://github.com/memgraph/memgraph/pull/1425)
- All the memory allocated inside a procedure is tracked in order to notify when
  limits are reached. [#1443](https://github.com/memgraph/memgraph/pull/1443)
- [`date` and `localTime`](https://memgraph.com/docs/querying/functions#temporal-functions) can now be
  extracted from `LocalDateTime` data type.
  [#1381](https://github.com/memgraph/memgraph/pull/1381)

#### Bug fixes

- Query modules [`schema`](https://memgraph.com/docs/querying/functions#schema) and
  [`convert`](https://memgraph.com/docs/querying/functions#conversion-functions) that had been moved from
  MAGE library to Memgraph now load properly and work as expected.
  [#1490](https://github.com/memgraph/memgraph/pull/1490)
- Instead of crashing the database, using the
  [`exists()`](https://memgraph.com/docs/querying/functions#pattern-functions) function with any clause
  other than WHERE throws an exception.
  [#1392](https://github.com/memgraph/memgraph/pull/1392)
  [#1382](https://github.com/memgraph/memgraph/pull/1382)
- The [`MEMORY QUERY LIMIT`
  queries](https://memgraph.com/docs/fundamentals/storage-memory-usage#control-query-memory-usage) now
  respect the memory limit as expected.
  [#1468](https://github.com/memgraph/memgraph/pull/1468)
- Queries using indexed join no longer crash.
  [#1478](https://github.com/memgraph/memgraph/pull/1478)
- Indexes can now be loaded in parallel as expected.
  [#1479](https://github.com/memgraph/memgraph/pull/1479)
- The [`collect` function](https://memgraph.com/docs/querying/functions#aggregation-functions), when
  called on an empty list, now yields no results instead of returning `NULL`.
  [#1482](https://github.com/memgraph/memgraph/pull/1482)
- Using the `WHERE` and `IN` subclauses with lists now works as expected.
  [#1494](https://github.com/memgraph/memgraph/pull/1494)
- The hash join executes as expected when using WHERE subclause.
  [#1496](https://github.com/memgraph/memgraph/pull/1496)
- When encountering corrupted [snapshot and WAL
  files](https://memgraph.com/docs/configuration/data-durability-and-backup#durability-mechanisms) during
  recovery, Memgraph now deletes them.
  [#1385](https://github.com/memgraph/memgraph/pull/1385)
- The snapshot thread is now stopped when switching from
  `IN_MEMORY_TRANSACTIONAL` to `IN_MEMORY_ANALYTICAL` [storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#storage-modes). Before, the snapshot
  thread was left running in the ANALYTICAL mode although it was unnecessary.
  [#1385](https://github.com/memgraph/memgraph/pull/1385)
- When setting properties to all nodes at once using the `SET` subclause via
  client library, the number of set properties is returned as expected.
  [#1460](https://github.com/memgraph/memgraph/pull/1460)
- Passing a variable that contains a list to the [BFS
  algorithm](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#breadth-first-search)
  now returns a results instead of an exception.
  [#1380](https://github.com/memgraph/memgraph/pull/1380)
- Upon deleting data, garbage collector deallocates memory as expected.
  [#1471](https://github.com/memgraph/memgraph/pull/1471)
- If you call a procedure that works on multiple threads and they pass the
  memory allocation limit, an exception will be thrown. Until now RAM
  suffered in silence. [#1401](https://github.com/memgraph/memgraph/pull/1401)

### Lab v2.9.1 - Nov 15, 2023

#### Bug fixes

- Memory and disk usage are now presented in KiB, MiB, and GiB instead of KB,
  MB, and GB.
- Memory usage is not N/A anymore, it shows the correct value for Memgraph
  versions higher than `2.12.1`.

### Memgraph v2.12.0 - Nov 1, 2023

#### Upgrading to Memgraph v2.12.0

> **Info**
>
> - The [`SHOW STORAGE INFO;`](https://memgraph.com/docs/configuration/server-stats) query now returns memory
> allocations in B, KiB, MiB, GiB or TiB, rather than just B so if any of your
> code depends on this information be sure to make the appropriate adjustments.
>
> - Snapshot and WAL files versions have been updated. Snapshot and WAL files created
> with Memgraph v2.12.0 and higher won't be usable in previous versions of
> Memgraph (v2.11.1 and lower).
>
> - You no longer need the MAGE library to run `schema` procedures or
> `convert.str2object()` function.

#### New features and improvements

- Snapshot and WAL files version have been updated. Snapshot and WAL files
  created with Memgraph v2.12 and higher won't be usable in previous versions of
  Memgraph (v2.11.1 and lower).
  [#1281](https://github.com/memgraph/memgraph/pull/1281)
- Memgraph can now start the instance in any [storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#storage-modes), defined by the
  `--storage-mode` [configuration
  flag](https://memgraph.com/docs/configuration/configuration-settings#storage).
- The [`SHOW STORAGE INFO;`](https://memgraph.com/docs/configuration/server-stats) query now returns
  memory allocations in B, KiB, MiB, GiB or TiB, rather than just B to make the
  information easily
  readable.[#1366](https://github.com/memgraph/memgraph/pull/1366)
- [Client libraries](https://memgraph.com/docs/client-libraries) you can use to connect to Memgraph are
  now able to retry the queries in order to complete the transactions
  successfully. [#1361](https://github.com/memgraph/memgraph/pull/1361)
- Information about the graph after running the [`ANALYZE
  GRAPH;`](https://memgraph.com/docs/fundamentals/indexes#analyze-graph) is now persistent
  between instance reruns.
  [#1281](https://github.com/memgraph/memgraph/pull/1281)
- Now, you can employ [index
  hints](https://memgraph.com/docs/fundamentals/indexes#index-hinting) by adding the
  `USE INDEX` subclause to the `CREATE INDEX` query to fine-tune the planner
  behavior. [#1345](https://github.com/memgraph/memgraph/pull/1345)
- In certain situations, Memgraph returns a hint to create
  [indexes](https://memgraph.com/docs/fundamentals/indexes) to speed up read queries.
  [#1343](https://github.com/memgraph/memgraph/pull/1343)
- The `SHOW INDEX INFO;` now also returns the information about the number of
  assigned indexes. [#1229](https://github.com/memgraph/memgraph/pull/1229)
- Now, you can [register REPLICA
  instances](https://memgraph.com/docs/configuration/replication#registering-replica-instances) using DNS
  names. [#1323](https://github.com/memgraph/memgraph/pull/1323)
- You can now use [`keys()` and `values()` functions](https://memgraph.com/docs/querying/functions#maps)
  to get a list of keys and values of a map.
  [#1246](https://github.com/memgraph/memgraph/pull/1246)
- You can now use `schema` procedures to get information about nodes and
  relationships or `convert.str2object()` function to convert strings to objects
  using Python's `json.dumps` function.
  [#1384](https://github.com/memgraph/memgraph/pull/1384)
- In your [custom query modules](https://memgraph.com/docs/custom-query-modules), you can now remove the
  last element of a path in [C++ API](https://memgraph.com/docs/custom-query-modules/cpp) using
  `path.Pop()`, [C API](https://memgraph.com/docs/custom-query-modules/c) using `mgp_path_pop` and
  [Python API](https://memgraph.com/docs/custom-query-modules/python) using `path.pop()`.
  [#1249](https://github.com/memgraph/memgraph/pull/1249)
- [`KeyExists` function](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api#keyexists) that
  checks the existence of a key in a map has been added to the C++ API.
  [#1336](https://github.com/memgraph/memgraph/pull/1336)
- You can now change the relationship type using `mgp::Graph::ChangeType` in [C++
  API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api) and `mgp::graph_edge_change_type` in
  [C API](https://memgraph.com/docs/custom-query-modules/c/c-api).
  [#1364](https://github.com/memgraph/memgraph/pull/1364)
- Reading of nodes [stored on
  disk](https://memgraph.com/docs/fundamentals/storage-memory-usage#storage-modes) is now 2x faster.
  [#1266](https://github.com/memgraph/memgraph/pull/1266)
- Searching for relationships stored on disk is now faster due to the ability to
  expand from a single node, rather than searching through all relationships.
  [#1335](https://github.com/memgraph/memgraph/pull/1335)
- Due to several smaller improvements, Memgraph uses less memory and the garbage
  collection is more efficient.
  [#1387](https://github.com/memgraph/memgraph/pull/1387)
- The [query inspection](https://memgraph.com/docs/querying/query-plan)
  report using `EXPLAIN` and `PROFILE` clauses now provides information about the
  operands of the `Filter` operator.
  [#1265](https://github.com/memgraph/memgraph/pull/1265)
- Queries that have multiple `MATCH` clauses and the results are joined in a
  `WHERE` clause now execute more efficiently, due to the to the new [`HashJoin`
  operator](https://memgraph.com/docs/querying/query-plan#query-plan-operators).
  [#1226](https://github.com/memgraph/memgraph/pull/1226)
- The new [Cartesian product
  operator](https://memgraph.com/docs/querying/query-plan#cartesian-product) optimizes
  queries that match nodes without relationship.
  [#1193](https://github.com/memgraph/memgraph/pull/1193)
- Filtering and producing expressions in WITH and FILTER clauses have been
  optimized by adding cache.
  [#1432](https://github.com/memgraph/memgraph/pull/1432)
- The verification of [unique and existence
  constraints](https://memgraph.com/docs/fundamentals/constraints) has been optimized. Constraints are
  now verified only when updating node labels or properties, as updates on
  relationships do not influence constraints.
  [#1341](https://github.com/memgraph/memgraph/pull/1341)

 
#### Bug fixes

- The `DISTINCT` clause no longer influences other aggregates, that is, shows
  distinct values as expected. [#1235](https://github.com/memgraph/memgraph/pull/1235)
- Any attempt to change the isolation level in in-memory analytical and on-disk
  storage mode will throw an exception. [#1367](https://github.com/memgraph/memgraph/pull/1367)
- A bug that would cause the memory limit not to be honored has been fixed.
  [#1250](https://github.com/memgraph/memgraph/pull/1250)
- The `SHOW CONFIG;` query now shows the correct values for the configuration
  settings, even if they were changed during runtime.
  [#1278](https://github.com/memgraph/memgraph/pull/1278)
- Procedure names of the key:value pairs mapped in a JSON file no longer need to
  be identical to be correctly mapped.
  [#1252](https://github.com/memgraph/memgraph/pull/1252)
- There are no longer data discrepancies between a REPLICA instance which is
  syncin with a new MAIN instance that became MAIN by being upgraded from the
  REPLICA role. [#1370](https://github.com/memgraph/memgraph/pull/1370)
- A bug has been fixed which would shut down Memgraph when fetching data from SQL
  database. [#1299](https://github.com/memgraph/memgraph/pull/1299)
- When setting a memory limit using [`--memory-limit` config
  flag](https://memgraph.com/docs/configuration/configuration-settings#other), Memgraph now stops the
  transactions exceeding the limit as expected.
  [#1250](https://github.com/memgraph/memgraph/pull/1250)
  [#1340](https://github.com/memgraph/memgraph/pull/1340)
- Memgraph now throws an exception when its out of memory, instead of not
  reporting an issue. [#1379](https://github.com/memgraph/memgraph/pull/1379)
- Running a `DELETE DETACH` query on a non-existing relationship will no longer
  crash the database. [#1335](https://github.com/memgraph/memgraph/pull/1355)

### MAGE v1.12.0 - Nov 1, 2023

- With the new [`algo`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/algo) query
  module you can run A* algorithm or search for all simple paths.
  [#353](https://github.com/memgraph/mage/pull/353)
  [#371](https://github.com/memgraph/mage/pull/371)
  [#356](https://github.com/memgraph/mage/pull/356)
- With the new [`merge`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/merge) query
  module you can merge or create nodes and relationships as per specified
  conditions, with more precision and coherence.
  [#390](https://github.com/memgraph/mage/pull/390)
- The [`node`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/node) query module now
  has procedures that return the in and out degrees of a node.
  [#369](https://github.com/memgraph/mage/pull/369)
- The [`path`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/path) query module now
  has functions that convert a given path into a list with a
  node-relationship-node order, as well as combine two paths into one and return
  them as a list in the node-relationship-node order.
  [#394](https://github.com/memgraph/mage/pull/394)
- The [`refactor`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/refactor) query
  module now has a procedure you can use to rename a property of a relationship,
  delete certain nodes in a path from the graph and reconnect the path and
  normalize properties to boolean values.
  [#376](https://github.com/memgraph/mage/pull/376)
- You no longer need the MAGE library to run `schema` procedures and get
  information about nodes and relationships or `convert.str2object()` function
  to convert strings to objects using Python's `json.dumps` function.
  [#1384](https://github.com/memgraph/memgraph/pull/1384)

### Lab v2.9.0 - Oct 31, 2023

- You can now run parameterized Cypher queries with query parameters defined using
  JSON format.
- Memgraph Lab now indicates there's a new version available for download and it
  allows you to check the release notes.
- You can expand and collapse nodes by double-clicking on them in graph
  view.
- The performance of the expand node query has been improved.

### Lab v2.8.4 - Oct 12, 2023

#### Improvements

- Increase the font size of the Cypher, GSS, and module editor using keyboard
  shortcuts `CMD/CTRL` + `=` or `CMD/CTRL` + `-`.

#### Bug fixes

- Sections to copy and download results within query collections now close as
  expected.
- Desktop version od Memgraph Lab no longer disconnects from Memgraph database
  instance within the first 3 seconds.
- Query module procedures and functions link to the correct documentation link.

### Lab v2.8.3 - Sep 29, 2023

#### Improvements

- The status bar now shows Memgraph's current [storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#storage-modes): transactional,
  analytical, or on-disk.

#### Bug fixes

- The reconnection screen now shows the correct Memgraph version and storage
  details, such as RAM and disk size.
- All the help resources bound to the new Memgraph docs are now available.
- Query result that would return a map with `type: "node"` no longer returns an
  error.
- The window in the *Collections* section showing details about query runtimes now
  closes as expected.
- Query error messages no longer show the redundant *”Query failed:”* prefix.
- The `docker run` command visible on the connection screen can now be correctly
  copied and pasted.

### MAGE v1.11.1 - Sep 29, 2023

#### Improvements

- Input and output parameter names of
  [collections](https://memgraph.com/docs/advanced-algorithms/available-algorithms/collections),
  [create](https://memgraph.com/docs/advanced-algorithms/available-algorithms/create),
  [label](https://memgraph.com/docs/advanced-algorithms/available-algorithms/label),
  [map](https://memgraph.com/docs/advanced-algorithms/available-algorithms/map),
  [refactor](https://memgraph.com/docs/advanced-algorithms/available-algorithms/refactor) and
  [schema](https://memgraph.com/docs/advanced-algorithms/available-algorithms/schema) query modules have
  been changed to align with Memgraph's GraphQL mappings. Also some procedures
  are now functions. Functions can be called inline, while procedures are called
  using the `CALL` clause. [#361](https://github.com/memgraph/mage/pull/361)
  [#362](https://github.com/memgraph/mage/pull/362)

### Memgraph v2.11.0 - Sep 13, 2023

> **Warning**
>
> `Node Properties()` and `Relationship
> Properties()` getter from the [C++ API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api)
> changed return type from `std::map` to `std::unordered_map` to speed up
> execution time. Check your C++ query modules for possible code changes.

#### New features and improvements

- `Node Properties()` and `Relationship Properties()` getter from the [C++
  API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api) changed return type from `std::map` to
  `std::unordered_map` to speed up execution time. This is a breaking change so
  check your C++ query modules for possible code changes.
  [#1131](https://github.com/memgraph/memgraph/pull/1131)

- The import of relationships in [on-disk storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#on-disk-transactional-storage-mode) can be
  improved by activating the `EDGE IMPORT MODE ACTIVE;`. When `EDGE IMPORT MODE`
  is disabled, the engine can import around 180 relationships per second. When
  active, the engine can import around 7500 relationships per second.
  [#1157](https://github.com/memgraph/memgraph/pull/1157).
- The following [configurational
  settings](https://memgraph.com/docs/configuration/configuration-settings) can now be changed during
  runtime: server name, query execution timeout, log level and the option to log
  to `stderr`. Use the following queries to set the configuration:
```
SET DATABASE SETTING 'server.name' TO 'new-name';
SET DATABASE SETTING 'query.timeout' TO '100';
SET DATABASE SETTING 'log.level' TO 'TRACE';
SET DATABASE SETTING 'log.to_stderr' TO 'true`;
```
  [#1183](https://github.com/memgraph/memgraph/pull/1183)
- The default value of
  [`--bolt-server-name-for-init`](https://memgraph.com/docs/configuration/configuration-settings#bolt) is
  now `Neo4j/v5.11.0 compatible graph database server - Memgraph`.
  [#1183](https://github.com/memgraph/memgraph/pull/1183)
- During recovery from a snapshot, the recovery of each graph object or property
  is no longer logged in the TRACE setting. The log now only indicates the
  recovery progress. [#1054](https://github.com/memgraph/memgraph/pull/1054)
- There have been various updates that improved performance:
  - Updating indexes and constraints has been streamlined, significantly improving
    execution time for everybody making heavy use of them.
    [#1159](https://github.com/memgraph/memgraph/pull/1159) [#1142](https://github.com/memgraph/memgraph/pull/1142)
  - Queries that build maps with multiple same-variable property lookups have been
    optimized. [#1168](https://github.com/memgraph/memgraph/pull/1168)
  - The batch update of properties improves performance when setting a large
    number of properties, as in this example:
```
FOREACH (i in range(0, 1000000) | CREATE (n:Label {id:i}));

CREATE INDEX ON :Label(id);

FOREACH (i IN range(0, 1000000, 3) | MERGE (n:Label {id:i}) SET n += {prop2:"a1", prop3:"b2", prop4:"c3", prop5:"d4", prop6:"e5", prop7:"f6", prop8:"g7", prop9:"h8", prop10:"i9", prop11:"j10 q"});
```
    [#1115](https://github.com/memgraph/memgraph/pull/1115)
  - Setting properties is also improved by caching mappings of property name to
    internal property `id`. [#1147](https://github.com/memgraph/memgraph/pull/1147)
  - Performance has been improved for concurrent operations contending on the same
    node. [#1187](https://github.com/memgraph/memgraph/pull/1187)
  - When a query is executing in many iterations over the graph entities, the
    performance has been improved by 100% due to faster scanning of nodes, for example:
```
UNWIND RANGE (1, 500) AS i CREATE ();
MATCH (),(),() RETURN COUNT(*);
```
    [#1127](https://github.com/memgraph/memgraph/pull/1227)
  - The query engine is more performant as at all times it is scanning and
    expanding nodes instead of scanning both source and destination nodes and
    then expanding to the relationship between them.
    [#1085](https://github.com/memgraph/memgraph/pull/1085)
  - The expansion of node is no longer only done from left to right. Depending on
    the previous executions and how many relationships needed to be check on
    MERGE, the engine knows the fewest vector necessary to expand.
    [#1110](https://github.com/memgraph/memgraph/pull/1110)
  - The performance of `DETACH DELETE` query has been improved.
  [#1078](https://github.com/memgraph/memgraph/pull/1078)
- When developing [custom query modules](https://memgraph.com/docs/custom-query-modules), node and
  relationship properties can now be changed in bulk, rather than one by one.
  [#1131](https://github.com/memgraph/memgraph/pull/1131)
- Changes to [the C++ API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api):
  - Users can now call `ToString()` method on `mgp::Value` and Memgraph's data
    types when writing query modules using C++ API.
    [#1140](https://github.com/memgraph/memgraph/pull/1140)
  - Trigger functions can now access deleted vertices from deleted edge when
    processed. [#1209](https://github.com/memgraph/memgraph/pull/1209)
  - When developing query modules using C++ API, you can now get the `In` and
    `Out` degrees of a node in O(1) time complexity. [#1217](https://github.com/memgraph/memgraph/pull/1217)
  - The C++ API now enables you to change relationship start (from) and end (to)
    nodes with `mgp::Graph.SetFrom` and `mgp::Graph.SetTo` methods.
- `SHOW INDEX INFO` now displays index information in alphabetic order for
  easier orientation. [#1178](https://github.com/memgraph/memgraph/pull/1178)
- The `PROFILE` query now generates a table with operators in the same order as
  in the plan constructed with the `EXPLAIN` query.
  [#1024](https://github.com/memgraph/memgraph/pull/1204)
- Certain queries execute faster due to the usage of custom allocators for
  [Delta
  objects](https://memgraph.com/docs/configuration/data-durability-and-backup#write-ahead-logging).
  [#1129](https://github.com/memgraph/memgraph/pull/1129)

#### Bug fixes

- Setting a property of a node labeled `null` will no longer crash the database,
  but return an exception. [#1175](https://github.com/memgraph/memgraph/pull/1175)

### MAGE v1.11.0. - Sep 13, 2023

#### New features and improvements

- The [`export_util`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/export_util) has
  been expanded and now allows exporting query results, as well as graph nodes
  and relationships to a CSV file format. [#325](https://github.com/memgraph/mage/pull/325)
- With the
  [`export_util`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/export_util) you can
  also export the graph to the Cypher query language in the specified file or
  stream, or a graphML file. [#348](https://github.com/memgraph/mage/pull/348)
  [#327](https://github.com/memgraph/mage/pull/327)
- The [import_util](https://memgraph.com/docs/advanced-algorithms/available-algorithms/import_util) can
  now also import data from graphML files. [#327](https://github.com/memgraph/mage/pull/327)
- The new [`util_module`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/util_module)
  offers a range of functions for tasks such as validation and creating
  cryptographic hash values. This module serves as a practical resource for
  streamlining a variety of tasks related to database operations.
  [#311](https://github.com/memgraph/mage/pull/311)
- The [`node` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/node) provides a
  comprehensive toolkit for managing individual graph nodes, enabling creation,
  deletion, updating, merging, and more.
  [#315](https://github.com/memgraph/mage/pull/315)
- The [`nodes` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/nodes) provides
  a comprehensive toolkit for managing multiple graph nodes, enabling linking,
  updating, type deduction and more.
  [#312](https://github.com/memgraph/mage/pull/312)
- By enabling more precise and flexible creation of nodes and relationships
  within the graph structure, the [`create`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/create) enhances the ability
  of developers to model, manipulate, and query complex graph data.
  [#307](https://github.com/memgraph/mage/pull/307)
- The [`meta` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/meta) provides a
  set of procedures for generating metadata about the database.
  [#296](https://github.com/memgraph/mage/pull/296)
- The [`path` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/path) allows
  users to explore different paths, filter relationships, and nodes based on
  specific criteria, and achieve more complex path-related tasks that go beyond
  the capabilities of native Cypher - whether you're seeking all possible paths
  between two nodes, subgraphs that meet certain conditions, or various other
  path-oriented operations. [#313](https://github.com/memgraph/mage/pull/313)
- The [`refactor` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/refactor)
  provides utilities for changing nodes and relationships.
  [#342](https://github.com/memgraph/mage/pull/342)
- The [`xml_module`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/xml_module)
  enhances graph database capabilities by providing support for loading and
  parsing XML data. [#329](https://github.com/memgraph/mage/pull/329)
- The [`convert` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/convert) is
  used for conversion of one data type to another.
  [#326](https://github.com/memgraph/mage/pull/326)
- The [`schema` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/schema) offers
  procedures that allow you to interact with and retrieve information about the
  database schema. [#341](https://github.com/memgraph/mage/pull/341)
- The [`neighbors` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/neighbors)
  provides tools for users to interact with and query nodes that have direct
  relationships to a specified node, enabling an examination of the immediate
  connections within the graph and thereby gaining insights into the network
  structure and connectivity. [#308](https://github.com/memgraph/mage/pull/308)
- The [`temporal` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/temporal)
  enables more sophisticated manipulation, conversion, and calculation of date
  and time. These functions handle temporal (time-related) operations and offer
  extended capabilities compared to the
  [`date`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/date) functions.
  [#299](https://github.com/memgraph/mage/pull/299)

### Lab v2.8.2 - Sep 13, 2023

#### New features and improvements

- Prepared datasets are now loaded 20-30 times faster, and 150 times faster if the
  database is in the [in-memory analytical storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage#in-memory-analytical-storage-mode) on Memgraph > 2.9.
- You will be warned if your custom CYPHERL dataset import fails during the
  import process.
- After being generated, the graph schema is cached and you can always
  regenerate it.
- Run queries will now show the full roundtrip time spent on the query instead
  of just how much Memgraph spent on the query. You will still have the ability
  to see the segmentation and how much each component of query parsing,
  planning, and execution spent time on the query (also shows seconds, and
  minutes if needed, e.g. ##.## s).

#### Bug fixes

- Map is now shown as expected when applying the System default Graph Style
  Script.

### Memgraph v2.10.1 - Aug 22, 2023

#### New features and improvements

- Removing properties from relationships with `RemoveProperty()` function in
  [C++
  API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api).
  [#1156](https://github.com/memgraph/memgraph/pull/1156)
- Improved performance on batch update of properties, e.g., `MATCH (n) SET n += {prop1:1, prop2:2, ...};`
  [#1115](https://github.com/memgraph/memgraph/pull/1115)
- When working at a snapshot isolation level, if a certain graph object is
  changed multiple times as a part of one of many parallel transactions, the
  delta chain tracking its changes becomes very long. Other parallel transaction
  with snapshot insolation that started before the modifications and dependent
  on that object need to access and process that delta chain to get its correct
  state, which can be very expensive regarding CPU usage. Now, Memgraph is
  caching delta chains of 128 or more changes, which eliminates the need of
  processing the chain and improves performance. If you would benefit from
  changing at what length the delta chain should be cashed, adjust the value of
  the [`--delta-chain-cache-threshold` configuration
  flag](https://memgraph.com/docs/configuration/configuration-settings#other).
  [#1124](https://github.com/memgraph/memgraph/pull/1124).

#### Bug fixes

- The same query module can now be executed concurrently by different clients
  without issues. [#1158](https://github.com/memgraph/memgraph/pull/1158)

### MAGE v1.10.0 - Aug 23, 2023

#### Features and improvements

- You can use the [`label`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/label) to check whether a
  node has the given label. [#304](https://github.com/memgraph/mage/pull/304)
- Time zones don’t have to be difficult to work with. The [`date`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/date) provides utilities for
  date and time operations. [#291](https://github.com/memgraph/mage/pull/291)
- The [`text` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/text) simplifies
  string concatenation and supports custom delimiters.
  [#287](https://github.com/memgraph/mage/pull/287)

#### Bug Fixes

- LLM prompts returned by the [`llm_util`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/llm_util) now list node and relationship
  properties in a consistent order.
  [#324](https://github.com/memgraph/mage/pull/324)

### Lab v2.8.1 - Aug 22, 2023

#### Bug fixes

- Quick connect will now try to connect to Memgraph on `127.0.0.1` instead of
  `localhost` because it is not always clear what `localhost` will be resolved
  to.
- Paths in query results now return a correct presentation instead of an empty
  path when there is only one node in the path.
- On hard refresh of the Lab application, you are now redirected to the correct
  connection screen.
- Logs no longer scroll automatically to new log records when you scroll up to
  check past log records.

### Memgraph v2.10.1 - Aug 22, 2023

#### Improvements and bug fixes

- Improved performance on batch update of properties, e.g.:

```cypher
MATCH (n) SET n += {prop1:1, prop2:2, ...};
```
  [#1115](https://github.com/memgraph/memgraph/pull/1115)
- Added a delta cache to improve query performance. This helps in situations of
  repeated reads of vertices which have many delta changes caused by another
  transaction while the current transaction is operating with snapshot isolation
  level and so needs to process those deltas. This can be tuned using
  `--delta-chain-cache-threshold`.
  [#1124](https://github.com/memgraph/memgraph/pull/1124)
- Concurrent access to the same query module had a race-condition on the
  pointer that was used to handle the custom memory management. A mapping has
  been added that keeps the information about what thread used what pointer to
  handle the memory resources, which should be fine since the respected query
  executions are running on a dedicated thread. Access to the mapping itself is
  thread-safe. A simple `RWLock` has been implemented here, as we shouldn't
  include `memgraph::utils` from this header and a traditional mutex might be
  overkill. A simple RAII wrapper for the mapping container has been also added
  for simpler client-side use.
  [#1158](https://github.com/memgraph/memgraph/pull/1158)

### Lab v2.8.0 - Aug 04, 2023

#### New features and improvements

- You can now save and edit connection details so you don't have to type them in
  on every reconnect.
- A list of recent connections is now shown in the sidebar and you can use it to
  quickly connect to any of the recent successful connections.
- When you click on a node or relationship in the Graph View, the object
  properties will be shown in a sidebar next to the Graph View rather then on
  the canvas next to the object.
- Nodes can be expanded by a double-click.
- As Memgraph now supports multi-tenant architecture, you can see the database
  you are currently working in the status bar with an option to switch to another
  database.
- Along with the number of nodes and relationships in the database, the status
  bar now also shows the number of indexes, constraints, and triggers.
- Now you can recenter, zoom in and zoom out the graph view by clicking on the
  action buttons in the bottom right corner of the Graph Results view.
- Now you can copy query results to the clipboard in JSON, CSV, or TSV format.
- You can start a name of a variable in Graph Style Script with an underscore
  `_` which is pretty handy for defining local variables in `@NodeStyle` and
  `@EdgeStyle` directives.

#### Bug fixes

- Setting up the `z-index` for edges in Graph Style Script now works as expected.
- Large integer numbers in query table results are now shown correctly, e.g. it
  shows `10,000,000,000,000,001` for the query `RETURN 10000000000000000 + 1`
  instead of incorrect result `10000000000000000`.
- Results now correctly show "No results" when there are no results in the query
  response, instead of a previously fetched result.
- The Lab web application is no longer unexpectedly crashing when using the
  Memgraph Platform Docker image.

### MAGE v1.9.0 - Aug 03, 2023

#### Features and improvements

- With the new [`collections` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/collections) you can
  filter, sort and modify lists within Cypher queries.
  [#284](https://github.com/memgraph/mage/pull/284)
- The new [`map` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/map) enables manipulating
  collections of key-value pairs, and consequently advanced data operations
  within a graph database context.
  [#282](https://github.com/memgraph/mage/pull/282)

### Memgraph v2.10.0 - Aug 2, 2023

#### New features and improvements

- The new [multi-tenant support](https://memgraph.com/docs/configuration/multi-tenancy) available in
  the Enterprise Edition of Memgraph enables you to manage multiple isolated
  databases within a single instance. The primary objective is to facilitate
  efficient resource isolation, maintain data integrity, and manage access for
  different clients. [#952](https://github.com/memgraph/memgraph/pull/952)
- The configuration flag `storage-recover-on-startup` has been deprecated and
  replaced with `data-recovery-on-startup`to support multi-tenancy.
- If you want to replace procedure names your application calls without changing
  the application code, you can define the mapping of the old and new procedure
  names in a JSON file, then set the path to the files as the value of the
  `query-callable-mappings-path` [configuration
  flag](https://memgraph.com/docs/configuration/configuration-settings).
  [#1018](https://github.com/memgraph/memgraph/pull/1018)
- The [C++
  API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api)
  for writing custom query modules now enables:
  - inserting `mgp::Any` datatype into Record. [#1094](https://github.com/memgraph/memgraph/pull/1094)
  - comparing two `mgp::Value` variables with the `<` operator. [#1090](https://github.com/memgraph/memgraph/pull/1090)
  - printing the type of `mgp::Type` enumeration using the `<<` operator. For
  example, if you have a `mgp::List` list, `cout<< list <<endl` will output
  `"list"`. [#1080](https://github.com/memgraph/memgraph/pull/1080)
  - printing `mgp::Value` variables (except `mgp::Path`, `mgp::List` and
    `mgp::Map` types) using the `<<` operator.
    [#1127](https://github.com/memgraph/memgraph/pull/1127)
  - using the `mgp::Value` variables and all its subtypes (`mgp::Map`,
    `mgp::Path`, ...) inside hash structures such as `std::unordered_map` and
    `std::unordered_set`. [#1093](https://github.com/memgraph/memgraph/pull/1093)
  - deleting and updating map elements with `mgp::Map.Update(key, &value)`,
    `mgp::Map.Update(key, &&value)` and `mgp::Map.Erase(key)` functions.
    [#1103](https://github.com/memgraph/memgraph/pull/1103)
  - removing properties from nodes with `RemoveProperty()` function and labels
    with `RemoveLabel()` function.
    [#1128](https://github.com/memgraph/memgraph/pull/1128)
    [#1126](https://github.com/memgraph/memgraph/pull/1126)

  Also, the `mgp::Value` wrapper for Memgraph's data types has been extended to
  return subtypes which are modifiable (non-const).
  [#1099](https://github.com/memgraph/memgraph/pull/1099)
- The [C
  API](https://memgraph.com/docs/custom-query-modules/c/c-api)
  for writing custom query modules now enables:
    - deleting and updating map elements with `mgp_map_update(map, key, value)`
  and `mgp_map_erase(map, key)` functions. [#1103](https://github.com/memgraph/memgraph/pull/1103)
    - removing labels from nodes with `RemoveLabel()` function.
      [#1126](https://github.com/memgraph/memgraph/pull/1126)
- Memgraph supports transaction timeouts defined by the Bolt protocol if the
  connection to the database is established via the [JavaScript
  client](https://memgraph.com/docs/client-libraries/javascript).
  [#1046](https://github.com/memgraph/memgraph/pull/1046)
- Queries exploring all shortest paths now use considerably less memory without
  significant performance deterioration.
  [#981](https://github.com/memgraph/memgraph/pull/981)
- Users with fine grained privileges will experience better performance when
  running queries due to user information cashing. Any changes to the user
  privileges will be ignored until the next login.
  [#1109](https://github.com/memgraph/memgraph/pull/1109)

#### Bug fixes

- Connection with Bolt v5.2 now works as expected when returning a path as a
  result. [#1108](https://github.com/memgraph/memgraph/pull/1108)
- Serializing vertex and edge properties to RocksDB now works as expected even
  when the serialization buffer is exactly 15B.
  [#1111](https://github.com/memgraph/memgraph/pull/1111)
- Users created in the Community Edition remain valid after the instance is
  upgraded to an Enterprise Edition. [#1067](https://github.com/memgraph/memgraph/pull/1067)

### Memgraph v2.9.0 - Jul 21, 2023

> **Warning**
>
> Memgraph 2.9 introduced a new configuration flag
> `--replication-restore-state-on-startup` which is `false` by default.
>
> If you want instances to remember their role and configuration in a replication
> cluster upon restart, the `--replication-restore-state-on-startup` needs to be
> set to `true` when first initializing the instances and remain `true` throughout
> the instances' lifetime.
>
> When reinstating a cluster it is advised to first initialize the MAIN
> instance, then the REPLICA instances.

#### New features and improvements

- The new [`ON_DISK_TRANSACTIONAL` storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage) allows you to store data on disk
  rather than in-memory. Check the implementation and implications in the
  reference guide. [#850](https://github.com/memgraph/memgraph/pull/850)
- Memgraph now works with all Bolt v5.2 drivers.
  [#938](https://github.com/memgraph/memgraph/pull/938)
- The [LOAD CSV clause](https://memgraph.com/docs/data-migration/csv) has several new improvements:
  - You can now import data from web-hosted CSV files by passing the URL as a
    file location. You can also import files compressed with `gzip` or `bzip2`
    algorithms. [#1027](https://github.com/memgraph/memgraph/pull/1027)
  - To speed up the execution of the `LOAD CSV` clause, you can add `MATCH` and
    `MERGE` entities prior to reading the rows from a CSV file. But, the `MATCH`
    or `MERGE` clause has to return just one row or Memgraph will throw an
    exception. [#916](https://github.com/memgraph/memgraph/pull/916)
  - If a certain sequence of characters in a CSV file needs to be imported as
    null, you can now specify them with the NULLIF option of the LOAD CSV
    clause. [#914](https://github.com/memgraph/memgraph/pull/914)
- You can now use `mgp::Type::Any` while developing a custom query procedure
  with [the C++
  API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api)
  to specify that the argument of the procedure can be of any type.
  [#982](https://github.com/memgraph/memgraph/pull/982)
- When you need to differentiate transactions, you can now define and pass
  [transaction metadata](https://memgraph.com/docs/fundamentals/transactions) via the client and check it in Memgraph by running the
  `SHOW TRANSACTIONS;` query.
  [#945](https://github.com/memgraph/memgraph/pull/945)
- You can now create [custom batch procedures in Python and
  C++](https://memgraph.com/docs/custom-query-modules)
  that process data in batches, thus consuming less memory.
[#964](https://github.com/memgraph/memgraph/pull/964)
- The [`ANALYZE GRAPH;` query](https://memgraph.com/docs/fundamentals/indexes#analyze-graph) now includes information about the degree of all
  nodes to enhance the MERGE optimizations on supernodes.
  [#1026](https://github.com/memgraph/memgraph/pull/1026)
- The `--replication-restore-state-on-startup` configuration flag allows you to
  define whether instances in the [replication
  cluster](https://memgraph.com/docs/configuration/replication) will regain their roles upon restart
  (`true`) or restart as disconnected "blank" MAIN instances (default setting
  `false`). This flag resolved the unwanted behavior of restarted REPLICA
  instances disconnecting from the cluster, but it also needs to be introduced
  to MAIN instances so they are not disconnected from the cluster upon restart.
  [#791](https://github.com/memgraph/memgraph/pull/791)

#### Bug fixes

- `init-file` and `init-data-file` configuration flags allow the execution of
  queries from a CYPHERL file, prior to or immediately after the Bolt server
  starts and are now possible to configure in the Community Edition as well.
  [#850](https://github.com/memgraph/memgraph/pull/850)
- The IN_MEMORY_ANALYTICAL storage mode now deallocates memory as expected and
  no longer consumes memory excessively. [#1025](https://github.com/memgraph/memgraph/pull/1025)
- When no values are returned from a map, a null is returned instead of an
  exception occurring. [#1039](https://github.com/memgraph/memgraph/pull/1039)

### MAGE v1.8.0 - Jul 21, 2023

#### Features and improvements

- With the [`llm_util` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/llm_util) you can generate a graph schema in a format best
  suited for large language models (LLMs).
  [#225](https://github.com/memgraph/mage/pull/225)
- When executing complex queries, the [`periodic`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/periodic) allows batching results from one query
  into another to improve execution time.
  [#221](https://github.com/memgraph/mage/pull/221)
- The [`do`
  module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/do) which allows the
  execution of different queries depending on certain conditions being met, has
  been rewritten from Python to C++ to improve performance and it can also
  periodically iterates. [#222](https://github.com/memgraph/mage/pull/222)
- The [`migrate` module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/migrate) has the option to get data from MySQL, SQL server, or
  Oracle DB to migrate it to Memgraph.
  [#209](https://github.com/memgraph/mage/pull/209)

### Lab v2.7.1 - Jul 05, 2023

#### Improvements

- System Default style has been renamed to System Style.
- If you run a query that has errors in the Graph Style Script code, you can
  decide to run it using the System Style.

#### Bug fixes

- Bug that would allow multiple styles to be the default has been fixed.
- The System Default Style now has `background-color` set to white.
- Queries selected in the Query Editor now execute as expected.
- Creating and editing a query module as well as selecting a transformation
  module in the Streams section now work as expected.
- All links are now linked with appropriate external resources.
- The pop-up window in the Run History that allows rerunning the query now
  closes once an option is selected.

### Lab v2.7.0 - Jun 28, 2023

#### What's new

- Now you can adjust the following settings:
  - Code completion and automatic graph rendering limits
  - The capacity of run history and its clearing
  - The limit for visible logs

  ![](https://memgraph.com/docs/pages/release-notes/lab-270/settings.png)
- The new interfaces for managing saved styles enables searching and changing
  the default style in the Lab. The saved styles now also have a preview.
  ![](https://memgraph.com/docs/pages/release-notes/lab-270/styles.png)
- The run history now also tracks changes to the query, style, or both. You can
  also filter out records to show All (both queries runs and applied styles),
  Query history (only query runs), and Style history (only applied style
  changes). You can expand both the query and style to see the full Cypher or
  GSS code.
  ![](https://memgraph.com/docs/pages/release-notes/lab-270/run_history.png)
  ![](https://memgraph.com/docs/pages/release-notes/lab-270/run_history_expand.png)
  ![](https://memgraph.com/docs/pages/release-notes/lab-270/run_history_queries.png)
- Queries inside a collection can be expanded and collapsed by clicking on their
  name.
- When testing and trying out different functions in the GSS you can use
  single-line (`// comment`) and multi-line (`/* comment */`) comments in the
  GSS code editor without losing previous state.
- Change the canvas color of the graph view with the new property
  `background-color` in [`@ViewStyle`](https://memgraph.com/docs/data-visualization/graph-style-script/directive-properties).
- Change the stack order of how nodes and edges are rendered in the graph view
  with the property `z-index` in
  [`@NodeStyle`](https://memgraph.com/docs/data-visualization/graph-style-script/directive-properties) and
  [`@EdgeStyle`](https://memgraph.com/docs/data-visualization/graph-style-script/directive-properties) directives. It works
  the same as the CSS z-index property.
- Set up transparent colors with the [new GSS
  functions](https://memgraph.com/docs/data-visualization/graph-style-script/built-in-elements) `RGBA` and `HSLA`. You can also get
  the transparency value with the function `Alpha`.
- [New functions](https://memgraph.com/docs/data-visualization/graph-style-script/built-in-elements) allow more customizations: `Sort`, `Coalesce`, `Reverse`,
  `IsMap`, `AsMap`, `Execute`, `Get`, `Set`, `Del`, `MapKeys`, `MapValues`,
  `AsIterator`, `IsIterator`, `Next`
- [Global and local
    variables](https://memgraph.com/docs/data-visualization/graph-style-script#caching-results-for-faster-performance)
    make developing new styles easier:
  - Variable graph is now available outside `@NodeStyle` and `@EdgeStyle`
    context
  - Local variables can be defined with `Define` within `@NodeStyle` and
    `@EdgeStyle` context
- Memgraph Lab is now packaged as an RPM package and arm64 (M1 chip) for MacOS.

#### Bug fixes

- Running a selected part of the Cypher query would just place that selected
  part in the run history. Now, the full query will be saved in the run history,
  and on its run, only the selected part will be executed again.
- Rows that would hide when scrolling the data results view now preview as expected.
- The System Default style colors all the nodes with the same label with a
  unique color.
- When showing a graph view on a map, you will no longer see a progress
  percentage which is unnecessary as each node has a fixed and known position
  due to its latitude and longitude values.
- All tables across the Lab are responsive as expected.
- Layouts no longer cause memory leaks and work as expected.
- By fixing a bug, you can now successfully connect to Memgraph using a hostname
  that contains numbers in the top-level domain.
- Markdown lists in query descriptions are indented as expected.

### Memgraph v2.8.0 - May 18, 2023

#### New features and improvements

- Data recovery is now up to 6x faster depending on the number of available
  cores, as [snapshot loading is distributed among several
  threads](https://memgraph.com/docs/configuration/data-durability-and-backup#snapshots).
  [#868](https://github.com/memgraph/memgraph/pull/868)
- During the recovery, indexes can also  be created using multiple threads, thus
  speeding up the process. [#882](https://github.com/memgraph/memgraph/pull/882)
- In the Enterprise Edition, Memgraph now [exposes system
  metrics](https://memgraph.com/docs/configuration/monitoring-server) via an HTTP endpoint, so
  you can get information about transactions, query latency and various other
  properties. [#940](https://github.com/memgraph/memgraph/pull/940)
- It’s now possible to use the [map projection
  syntax](https://memgraph.com/docs/fundamentals/data-types#maps) to create maps. Map
  projections are convenient for building maps based on existing values and they
  are used by data access tools like GraphQL.
  [#892](https://github.com/memgraph/memgraph/pull/892)
- You can now check if [the data directory](https://memgraph.com/docs/configuration/data-durability-and-backup) is
  (un)locked with the `DATA DIRECTORY LOCK STATUS;` query.
  [#933](https://github.com/memgraph/memgraph/pull/933)
- You can now check the current [storage
  mode](https://memgraph.com/docs/fundamentals/storage-memory-usage) and [isolation
  levels](https://memgraph.com/docs/fundamentals/transactions) by running the `SHOW STORAGE INFO;`
  query. [#883](https://github.com/memgraph/memgraph/pull/883)
- Check the suspected [build type of the Memgraph
  executable](https://memgraph.com/docs/configuration/server-stats#build-information) by
  running the `SHOW BUILD INFO;` query.
  [#894](https://github.com/memgraph/memgraph/pull/894)
- Performance has been improved by optimizing the deallocation of resources in
  Memgraph's custom `PoolResource` memory allocator.
  [#898](https://github.com/memgraph/memgraph/pull/898)

#### Bug fixes

- Running Python procedures now consume less memory.
  [#932](https://github.com/memgraph/memgraph/pull/932)
- Memory allocation in LOAD CSV queries has been optimized to avoid performance
  degradation. [#877](https://github.com/memgraph/memgraph/pull/877)
- Query profiles of the LOAD CSV queries now show the correct values of memory
  usage. [#885](https://github.com/memgraph/memgraph/pull/885)

### Lab v2.6.0 - Apr 20, 2023

#### What's new

- If you execute multiple Cypher queries, you can now view the result of each
  query instead of viewing just the last result.
  ![results](https://memgraph.com/docs/pages/release-notes/lab-260/results.png)
- Besides exporting query results to JSON, you can also export them to CSV and
  TSV file format.
  ![download](https://memgraph.com/docs/pages/release-notes/lab-260/download.png)
- If the dataset contains millions or billions nodes and relationships, their
  count in the status bar will be in the following format: X.XXM or X.XXB.
- Syntax of code blocks in the query collection description can now be
  highlighted by using one of the following language
  styles: `cypher`, `bash`, `python`, `css`, `c`, `cpp`, `json`, `sql`,
  and `yaml`. Check the examples of syntax highlighting in the [Markdown
  Guide](https://www.markdownguide.org/extended-syntax/#syntax-highlighting).
  ![markdown](https://memgraph.com/docs/pages/release-notes/lab-260/markdown.png)
- [New functions of the Graph Style Script
  language](https://memgraph.com/docs/data-visualization/graph-style-script/built-in-elements) used for customizing graph appearance
  are: `Reduce`, `Sum`, `Avg`, `Min`, `Max`, `IsArray`, `Hue`, `Saturation`,
  `Lightness`, `HSL`.
  ![gss](https://memgraph.com/docs/pages/release-notes/lab-260/gss.png)

#### Bug fixes

- The initial node count has been removed from the connection initialization so
  connecting to the Memgraph instance containing a huge number of nodes will no
  longer cause a timeout.
- Run History now logs queries as expected.
- When switching between the map view and the default view the graph view no
  longer becomes unresponsive.
- Using a new line character `\n` in the query module doesn’t result in a new
  line, but an explicit character `\n`.
- Viewing the code of multiple query modules in the split screen now
  works as expected.
- Notifications no longer mix with the Query Editor and Query Collections visual
  elements.
- Pressing CMD/CTRL + S will save a query within a query collection execution section as intended.
- The autosave in query collection is now triggered on every query run as it was
  intended.
- Running a selected portion of the query won’t remove the rest of the query
  from the query collection execution view.
- GSS `Blue` function was returning a wrong number. It is fixed now. `Lighter`
  and `Darker` functions now work correctly as well because they were depending
  on the output from the `Blue` function.

### Memgraph v2.7.0 - Apr 5, 2023

#### New features and improvements

- You can now choose between [two different storage modes](https://memgraph.com/docs/fundamentals/storage-memory-usage):
  - Transactional mode - the default database mode that favors
    strongly-consistent ACID transactions using WAL files and periodic
    snapshots, but requires more time and resources during data import and
    analysis.
  - Analytical mode - speeds up import and data analysis but offers no ACID
    guarantees besides manually created snapshots.
  Switch between modes within the session using the `STORAGE MODE
  IN_MEMORY_{TRANSACTIONAL|ANALYTICAL};` query. [#772](https://github.com/memgraph/memgraph/pull/772)
- You can now call [subqueries](https://memgraph.com/docs/querying/clauses/call) inside existing queries using the CALL clause.
  [#794](https://github.com/memgraph/memgraph/pull/794)
- When you want to filter data using properties that all have label:property
  indexes set, you can make Memgraph analyze the properties on all or several
  labels with the [`ANALYZE GRAPH;`
  query](https://memgraph.com/docs/fundamentals/indexes#analyze-graph). By calculating the
  distribution of property values, Memgraph will be able to select the optimal
  index for the query and it will execute faster.
  [#812](https://github.com/memgraph/memgraph/pull/812)
- If, for example, you are no longer interested in the results of the query you
  ran, or the procedure you built is running in an infinite loop, you can [stop
  the transaction](https://memgraph.com/docs/fundamentals/transactions#managing-transactions)
  with the `TERMINATE TRANSACTIONS tid;` query; Find out the transaction ID with
  `SHOW TRANSACTIONS;` query.
  [#790](https://github.com/memgraph/memgraph/pull/790)
- With the [new flag](https://memgraph.com/docs/configuration/configuration-settings#other)
  `password-encryption-algorithm` you can choose between `bcrypt`, `sha256`, and
  `sha256-multiple` encryption algorithms. SHA256 offers better performance
  compared to the more secure but less performant bcrypt.
  [#839](https://github.com/memgraph/memgraph/pull/839)
- Import using the [LOAD CSV clause](https://memgraph.com/docs/data-migration/csv) has
  been further improved by using a memory allocator which will reuse memory
  blocks allocated while processing the `LOAD CSV` query.
  [#825](https://github.com/memgraph/memgraph/pull/825)

#### Bug fixes

- The users who have [global visibility on the
  graph](https://memgraph.com/docs/configuration/security#label-based-access-control) will
  experience a slight improvement in performance regarding label-based access
  control as the engine no longer check privileges for each node.
  [#837](https://github.com/memgraph/memgraph/pull/837)
- The [All shortest paths
  algorithm](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#all-shortest-paths)
  now supports multiedges. [#832](https://github.com/memgraph/memgraph/pull/832)

### MAGE v1.7.0 - Apr 5, 2023

#### Features and improvements

- The [new conditional execution
module](https://memgraph.com/docs/advanced-algorithms/available-algorithms/do) lets you define
  conditions not expressible in Cypher and and use them to control query
  execution. [#203](https://github.com/memgraph/mage/pull/203)

### MAGE v1.6.1 - Mar 20, 2023

#### Features and improvements

- With the `export_util.csv_query()` procedure, you can export query results to a CSV file or as a stream. [#199](https://github.com/memgraph/mage/pull/199)
- Similarity algorithms (`jaccard`, `overlap` and `cosine`) have been rewritten in C++ to improve performance. [#196](https://github.com/memgraph/mage/pull/196)

### Lab v2.5.0 - Mar 17, 2023

#### What's New

* If there are several Cypher queries in the query editor you can select a single query and run
  it without commenting out all the other queries.

![run-selected-query-in-the-lab](https://memgraph.com/docs/pages/release-notes/lab-250/lab-run-selected-demo.gif)

* You can now open multiple query executions views side by side and compare query execution speed or results.

![multiple-query-executions-in-the-lab](https://memgraph.com/docs/pages/release-notes/lab-250/lab-multiple-editors-demo.gif)

* Query modules are now sorted alphabetically for easier and faster browsing. A search box has also been added to query modules with more
  than 5 procedures to help narrow them down (e.g. `nxalg` query module has [49 procedures](https://memgraph.com/docs/advanced-algorithms/available-algorithms/nxalg)).
* When rendering a graph with more than 3,500 nodes or 8,500 relationships, which might take considerable amount of time to preview, you will be
  asked if you want to proceed with the graph visualization or switch to the data view.
* Besides manually saving changes in the Cypher query and GSS style editor in the query collections section, they will also be saved
  automatically after each query run.
* Memgraph Lab will now notify you of any product updates and offer various tips and tricks for using the Memgraph ecosystem.

#### Bug Fixes

* Cypher code suggestions can now handle labels and properties of 250k nodes and 500k relationships, compared to the previous limit of
  100k nodes and 200k relationships.
* Multiple scrollable elements of the query collections was making scrolling difficult. Now you can focus on a particular element and
  scroll through it by clicking on it.
* Browser's back button is now working as expected when using Lab as a web application.
* Data in the query results, query modules and query run history tables now loads faster making the scrolling smoother and improving
  the user experience.
* Graph schema is now generated even if the database has no relationships.
* In-progress feedback when generating graph schema and exporting datasets for graphs with more than 10M nodes
  is now previewed as expected.
* A scrolling issue with expanded results in the Data view where you couldn't see a completely expanded row because the
  scroll would jump to the next row is now fixed.
* Dataset cards no longer spread apart when conducting a search.

### Memgraph v2.6.0 - Mar 07, 2023

#### Major features and improvements

- Importing speed using the LOAD CSV clause has been improved due to two changes:
  -  Performance improvement in accessing values from large arrays or maps with numerous properties. [#744](https://github.com/memgraph/memgraph/pull/774)
  -  Upon creating a large number of node or relationship properties, properties are stored in a property store all at once, instead of individually. [#788](https://github.com/memgraph/memgraph/pull/788)
- Newly implemented `exists()` function allows using patterns as part of the filtering clause. Check the [Cypher Manual](https://memgraph.com/docs/querying/clauses/where) for usage. [#818](https://github.com/memgraph/memgraph/pull/818)
- With the new [Python mock query module API](https://memgraph.com/docs/custom-query-modules/python/mock-python-api), you can now develop and test Python query modules for Memgraph without having to run a Memgraph instance. The mock API is compatible with the Python API and thus developed modules can be added to Memgraph as-is. [#757](https://github.com/memgraph/memgraph/pull/757)
- Memgraph now supports Fedora 36 and Ubuntu 22.04 for ARM. [#787](https://github.com/memgraph/memgraph/pull/787) [#810](https://github.com/memgraph/memgraph/pull/810)

#### Bug fixes

-  `torch` and `igraph` can no longer be removed from the `svs.modules` cache to avoid issues after reload. [#720](https://github.com/memgraph/memgraph/pull/720)
-  Newly created nodes now comply with the set label based authorization rules. [#755](https://github.com/memgraph/memgraph/pull/755)
-  Constructing LocalDateTime objects with invalid parameters doesn’t crash Memgraph anymore, but throws an informative exception. [#819](https://github.com/memgraph/memgraph/pull/819)
-  Error message warning about incompatible `epoch_id` between a MAIN and REPLICA instance has been improved. [#786](https://github.com/memgraph/memgraph/pull/786)

### MAGE v1.6.0 - Jan 30, 2023

#### Major Features and Improvements

- The `setup` script now halts if the build fails on C++ or Rust side. [#194](https://github.com/memgraph/mage/pull/194)
- With the [`meta_util.schema()` procedure](https://memgraph.com/docs/advanced-algorithms/available-algorithms/meta_util), you can generate a graph schema as a graph result. [#187](https://github.com/memgraph/mage/pull/187)
- The execution of the `single` method multiple times has been improved by rewriting [the distance calculator](https://memgraph.com/docs/advanced-algorithms/available-algorithms/distance_calculator) from Python to C++. [#191](https://github.com/memgraph/mage/pull/191)
- [Dynamic graph analytics](https://memgraph.com/docs/advanced-algorithms/available-algorithms/betweenness_centrality_online) have been ported to C++ to improve performance. [#182](https://github.com/memgraph/mage/pull/182)
- [New module `elastic_search_serialization`](https://memgraph.com/docs/advanced-algorithms/available-algorithms/elasticsearch_synchronization) enables developers to serialize Memgraph into Elasticsearch instance using basic authentication. [#170](https://github.com/memgraph/mage/pull/170)

### Memgraph v2.5.2 - Jan 26, 2023

#### Bug Fixes

- Variables can be used inside nested [FOREACH clauses](https://memgraph.com/docs/querying/clauses). [#725](https://github.com/memgraph/memgraph/pull/725)
- [FOREACH clause](https://memgraph.com/docs/querying/clauses) can now use indexes if needed (e.g. in case of MERGE). [#736](https://github.com/memgraph/memgraph/pull/736)
- [C++ API](https://memgraph.com/docs/custom-query-modules/cpp/cpp-api) now allows setting and getting node and relationship properties. [#732](https://github.com/memgraph/memgraph/pull/732)
- [OPTIONAL MATCH](https://memgraph.com/docs/querying/clauses/optional-match) can now use label property indexes that are referencing the previously matched variables. [#736](https://github.com/memgraph/memgraph/pull/736)
- Iterating over all relationships in a graph now works as expected, as well as checking whether the graph contains a given relationship. [#743](https://github.com/memgraph/memgraph/pull/743)
- Implementation of the [All Shortest Paths algorithm](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#all-shortest-paths) was fixed so the paths are no longer duplicated when the upper bound is used. [#737](https://github.com/memgraph/memgraph/pull/737)

### MAGE v1.5.1 - Jan 20, 2023

#### Major Features and Improvements

- The version of MemgraphDB that will be used in the Docker image has been updated to 2.5.1.
  [#193](https://github.com/memgraph/mage/pull/193)

### Memgraph v2.5.1 - Jan 19, 2023

#### Bug Fixes

- The LOAD CSV clause now uses less RAM to load a whole CSV file. Modification
  made to improve the LOAD CSV operation, also influenced high memory usage
  operations with objects such as lists and map. Modifying or accessing elements
  inside those objects now also uses less RAM.
  [#712](https://github.com/memgraph/memgraph/pull/712)
- The logic of the `read_write_type_checker` was corrected so queries now get
  the right `rw_type`, making the replication system work as expected.
  [#709](https://github.com/memgraph/memgraph/pull/709)
- Bolt protocol has been improved by adding the server-assigned query ID (`qid`)
  as part of the transactions' metadata.
  [#721](https://github.com/memgraph/memgraph/pull/721)
- Fixed a trigger bug that would cause an error if Memgraph is configured to run
  without properties on edges. As a result of the fiy, triggers are now working
  as expected when there are no properties on edges.
  [#717](https://github.com/memgraph/memgraph/pull/717)

### MAGE v1.5.0 - Dec 20, 2022

#### Major Features and Improvements

- Now you can find ancestors (all the nodes from which a path exists ) and descendants (all nodes to which a path exists) starting from a certain node, sort directed acyclic graph in a way that a node which appears before others is first, return a subgraph from nodes using `connect_nodes` method, and create relationships between nodes in a list using the `chain_nodes` method.
  [#180](https://github.com/memgraph/mage/pull/180)
- C++ API is now aligned with Memgraph 2.5
  [#184](https://github.com/memgraph/mage/pull/184)
- Graph Coloring no longer outputs strings but vertices and integers. This allows you to use the result of graph coloring directly in Memgraph Lab.
  [#177](https://github.com/memgraph/mage/pull/177)

#### Bug Fixes

- By enabling module reset, you can now train and evaluate the model without shutting down the database.
  Class labels can now start from 0 or negative numbers.
  The low limit of the early stopping flag no longer prematurely stops the training of the model while running the Node classification module.
  [#173](https://github.com/memgraph/mage/pull/173)

### Memgraph v2.5.0 - Dec 13, 2022

#### Major Features and Improvements

- `DISTINCT` operator can now be used with aggregation functions. Until now, if
  you wanted to use an aggregation function with distinct values, you had to
  write a query similar to this one `WITH DISTINCT n.prop as distinct_prop
  RETURN COUNT(distinct_prop)`. Now you can use the `DISTINCT` operator like in
  the following query, `RETURN COUNT(DISTINCT n.prop)`.
  [#654](https://github.com/memgraph/memgraph/pull/665)
- You can now create a user before the Bolt server starts using the environment
  variables `MEMGRAPH_USER` for the username, `MEMGRAPH_PASSWORD` for the
  password and `MEMGRAPH_PASSFILE` file that contains username and password for
  creating the user in the following format: `username:password`.
  [#696](https://github.com/memgraph/memgraph/pull/696)
- With the new configuration flag `init_file` you can execute queries from the
  CYPHERL file which need to be executed before the Bolt server starts and with
  the configuration flag `init_data_file` you can execute queries from the
  CYPHERL file immediately after the Bolt server starts.
  [#696](https://github.com/memgraph/memgraph/pull/696)

#### Bug Fixes

- Constructors and assignment operators in the C++ query modules API now work as
  expected, and the API type check in the `ValueNumeric` method now correctly
  recognizes numeric types.
  [#688](https://github.com/memgraph/memgraph/pull/688)
- Error message support (`SetErrorMessage`) has been added to query methods that
  use the MAGE C++ API. [#688](https://github.com/memgraph/memgraph/pull/688)
- The `EmptyResult` sink operator was added to the Memgraph's planner. This
means that results produced by a query `MATCH (n) SET n.test_prop = 2` will get
exhausted which was a problem in some Bolt clients implementations, e.g in
Golang's client. [#667](https://github.com/memgraph/memgraph/pull/667)
- Fixed Python submodules reloading when calling `CALL mg.load()` and `CALL
  mg.load_all()`. Before, only the Python module would be reloaded, but now all
  dependencies get reloaded as well. This includes Python's utility submodules
  and Python packages, which means that the environment with Python packages can
  be changed without turning off the database.
  [#653](https://github.com/memgraph/memgraph/pull/653)

### Lab v2.4.0 - Dec 2, 2022

#### What's New

* Memgraph Lab now supports manual transaction workflows you can construct using transaction commands `BEGIN`, `COMMIT`, and `ROLLBACK`.
* Cypher intellisense has been updated to suggest new Cypher features from Memgraph 2.4.0 such as:
  * Privileges for user-role authorization.
  * Commands and privileges for label-based authorization.
  * Manual transaction commands: `BEGIN`, `COMMIT`, `ROLLBACK`.
  * Checking configuration with `SHOW CONFIG`.
  * All shortest path algorithm `allShortest`.
  * Graph projection function `project`.
  * Additional query module signature that accepts a projected graph as an optional first argument.
* Graph results view will check for nodes and relationships in arrays and projected graphs. It simplifies
  the visualization of a projected graph or an array of nodes/relationships without using `UNWIND`.

#### Bug Fixes

* Once the table results view is selected, the results of the following query run will also preview in the table results view, instead of automatically switching to the graph view.
* Exploring a dataset's query collection now works as expected. It opens up a list of queries that can be used to explore the dataset.
* Failed queries from the rich collections now return a detailed error message.
* _Save code changes_ button in rich collections will now be enabled only if there are unsaved changes for the Cypher query
  or GSS.
* A bug that would only show the first node label instead of all node labels in the table results view has been fixed.

### MAGE v1.4.0 - Nov 15, 2022

#### Major Features and Improvements

- Implemented Link prediction with [DGL](https://www.dgl.ai/).
  [#160](https://github.com/memgraph/mage/pull/160)
- Implemented Node classification with PyTorch.
  [#161](https://github.com/memgraph/mage/pull/161)
- Added igraph support.
  [#150](https://github.com/memgraph/mage/pull/150)
- Added _k_-means embedding clustering algorithm.
  [#105](https://github.com/memgraph/mage/pull/105)
- Added better support for C++ API.
  [#174](https://github.com/memgraph/mage/pull/174)

#### Bug Fixes

- Enable module reset to be able to train and evaluate without shutting down database, enable working with class labels which don't start from 0, and fix potential early stopping due to low limit in the Node classification module.
  [#173](https://github.com/memgraph/mage/pull/173)

### Memgraph v2.4.2 - Nov 7, 2022

#### Bug Fixes

- Fixed a bug when calling `AllShortestPath` with `id` function.
  [#636](https://github.com/memgraph/memgraph/pull/636)
- Fixed bug when getting iterating over in-edges of a Node.
  [#582](https://github.com/memgraph/memgraph/pull/613)

### Lab v2.3.1 - Nov 4, 2022

#### Bug Fixes

* Writing a single-line comment in the Cypher code no longer results in an error.
* Having different map tiles (e.g. "light" map tile on one map view, but "dark" map tile on another map view) for multiple graph map views in the rich collection is enabled and works as expected.
* Graph rendering freeze when toggling the map view on/off during the graph rendering process has been fixed.
* All the information about nodes and edges on the graph schema is now previewed as expected.
* A bug that would mix query title and description when queries are reordered in the rich query collection has been fixed.
* A bug that would not reset the description field when adding a new query to the query collection has been fixed.
* Saving a new style now works as expected. The active style is saved, not the last applied one.

### Lab v2.3.0 - Oct 24, 2022

#### What's New

* Add new updates to the prepared datasets:
  * Add a search bar for searching and filtering datasets.
  * Add featured (highlighted) datasets.
  * Add rich collections with prepared queries, descriptions, and GSS for each dataset.
* Add new updates to the latest queries:
  * Change the name from "Latest queries" to "Run history" because it contains both queries and GSS changes.
  * Show GSS changes in the "Run history" section.
* Replace previous collections with "Rich collections":
  * Add more context to each collected query: title, markdown description, Cypher query, and GSS.
  * Add the ability to run multiple query executions within the query collection.
  * Add import and export functionality of a collection.
* Add a new version of GSS:
  * Add new GSS directive `@ViewStyle` to configure physics, link distance, repel force, and view type (`default` or `map`).
  * Add new GSS directive `@ViewStyle.Map` to configure map tiles for map view.
  * Add new GSS functions: `Slice`, `Split`, `Replace`, `Trim`, `Nodes`, `Edges`, `IsNumber`, `IsBoolean`, `IsString`, `IsNull`.
  * Add new GSS node properties `latitude` and `longitude` used to define the latitude and longitude of each node for the map view.
* Integrate graph visualization library `orb`.
* Add the ability to enable/disable map background view for nodes with geo information.
* Add the ability to connect to Neo4j, load datasets, and run Cypher queries.

#### Bug Fixes

* Fix map view to use latitude and longitude from GSS style instead of `lat` and `lng` node properties.
* Fix the default GSS to match new the map view configuration by checking `lat` and `lng` node properties.

### MAGE v1.3.2 - Oct 10, 2022

#### Major Features and Improvements

- Allowed restricting community detection to subgraphs.
  [#152](https://github.com/memgraph/mage/pull/152)
- Implemented the degree centrality algorithm.
  [#162](https://github.com/memgraph/mage/pull/162)
- Updated Memgraph version.
  [#171](https://github.com/memgraph/mage/pull/171)

#### Bug Fixes

- Dynamic betweenness centrality bugfix.
  [#147](https://github.com/memgraph/mage/pull/147)

### Memgraph v2.4.1 - Oct 7, 2022

#### Bug Fixes

- Fixed bug when getting EdgeType from Edge object or Label from Vertex object
  in query modules. [#582](https://github.com/memgraph/memgraph/pull/582)
- Fix a bug when changing role permissions for label based authorization, by
  passing user's instead of role's `fine_grained_access_handler`.
  [#579](https://github.com/memgraph/memgraph/pull/579)

### Memgraph v2.4.0 - Sep 15, 2022

#### Major Features and Improvements

- Add replica state to `SHOW REPLICAS` query.
  [#379](https://github.com/memgraph/memgraph/pull/379)
- Add `current_timestamp` and `number_of_timestamp_behind_master` to `SHOW
  REPLICAS` query. [#412](https://github.com/memgraph/memgraph/pull/412)
- Query `REGISTER REPLICA replica_name SYNC` no longer supports `TIMEOUT`
  parameter. To mimic the previous behavior of `REGISTER REPLICA replica_name
  SYNC WITH TIMEOUT 1`, one should use `REGISTER REPLICA replica_name ASYNC`
  instead. [#423](https://github.com/memgraph/memgraph/pull/423)
- Make behavior more [openCypher](http://opencypher.org/) compliant regarding
  checking against `NULL` values is `CASE` expressions.
  [#432](https://github.com/memgraph/memgraph/pull/432)
- Previously registered replicas are automatically registered on restart of
  Memgraph. [#415](https://github.com/memgraph/memgraph/pull/415)
- Add new command `SHOW CONFIG` that returns the configuration of the currently
  running Memgraph instance.
  [#459](https://github.com/memgraph/memgraph/pull/459)
- Extend the shortest paths functionality with [All Shortest
  Path](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#all-shortest-paths)
  query. [#409](https://github.com/memgraph/memgraph/pull/409)
- Extend the query modules C and Python API to enable logging on different
  levels. [#417](https://github.com/memgraph/memgraph/pull/417)
- Added C++ query modules API. Instead of using the C API call, C++ API calls
  significantly simplify the implementation of fast query modules.
  [#546](https://github.com/memgraph/memgraph/pull/546)
- [Enterprise] Added support for label-based authorization. In addition to
  clause-based authorization rules, each user can now be granted `NOTHING`,
  `READ`, `UPDATE`, or `CREATE_DELETE` permission on a given label or edge
  type. [#484](https://github.com/memgraph/memgraph/pull/484)
- New Cypher function `project()` creates a projected graph consisting of nodes
  and edges from aggregated paths. Any query module or algorithm can be now run
  on a subgraph, by passing the variable of the projected graph as the first
  argument of the query module procedure. [#535](https://github.com/memgraph/memgraph/pull/535)

#### Bug Fixes

- Added a check to ensure two replicas cannot be registered to an identical
  end-point. [#406](https://github.com/memgraph/memgraph/pull/406)
- `toString` function is now able to accept `Date`, `LocalTime`, `LocalDateTime`
  and `Duration` data types.
  [#429](https://github.com/memgraph/memgraph/pull/429)
- Aggregation functions now return the openCypher-compliant results on `null`
  input and display the correct behavior when grouped with other operators.
  [#448](https://github.com/memgraph/memgraph/pull/448)
- Corrected inconsistencies and incorrect behavior with regards to sync
  replicas. For more detail about the behavior, please check [Under the
  hood view on replication](https://memgraph.com/docs/configuration/replication).
  [#435](https://github.com/memgraph/memgraph/pull/435)
- Fixed handling `ROUTE` Bolt message. Memgraph didn't handle the fields of
  `ROUTE` message properly. Therefore the session might be stuck in a state
  where even the `RESET` message did not help. With this fix, sending a `RESET`
  message will properly reset the session.
  [#475](https://github.com/memgraph/memgraph/pull/475)

### Lab v2.2.1 - Aug 12, 2022

#### What's New

* Add improved and more precise progress when importing built-in datasets.
* Add an indicator for the total count of error log messages in the sidebar.
* Change the color scheme of code snippets for query modules.
* Add a help section when Lab's connection is reconnecting.
* Add breadcrumbs for the layout titles.

#### Bug Fixes

* Fix issues with query collections.
* Fix vertical layout usability when the help sidebar is opened.
* Fix various UI and UX issues across the application.
* Fix query results on the reconnected connection.

### Lab v2.2.0 - Jul 15, 2022

#### What's New

* Add a new table look and feel across the application: query results, the latest queries, modules, streams.
* Add a help section with relevant links, guides, and documentation search capability.
* Add test parameters (batch size, timeout) for testing stream transformation.
* Add new GSS functions: `Round`, `Floor`, and `Ceil`.

#### Bug Fixes

* Fix various issues in graph view, streams, and query collections.

### MAGE v1.3.1 - Jul 14, 2022

#### Major Features and Improvements

- Updated Memgraph version.
  [#154](https://github.com/memgraph/mage/pull/154)
- Introduced E2E group testing.
  [#145](https://github.com/memgraph/mage/pull/145)

### Lab v2.1.2 - Jun 21, 2022

#### What's New

* Add a dashboard and overview page for the better onboarding experience.
* Add environment variables for query, modules, and streams name length validator limits.
* Add logs connection status messages in the logs view.

#### Bug Fixes

* Fix several bugs with the stream configuration creation.
* Fix showing the logs when connected to Memgraph via an encrypted SSL connection.

### Memgraph v2.3.1 - Jun 23, 2022

#### Improvement

- Updated results return by [`CHECK
  STREAM`](https://memgraph.com/docs/data-streams#check-stream) query to group
  all queries/raw messages on single line per batch.
  [#394](https://github.com/memgraph/memgraph/pull/394)
- Add frequent replica ping. `main` instance checks state of the replicas with
  given frequency controller by `--replication-replica-check-delay-sec`. The
  check allows latest information about the state of each replica from `main`
  point of view. [#380](https://github.com/memgraph/memgraph/pull/380)
- Added `BATCH_LIMIT` and `TIMEOUT` options to [`START
  STREAM`](https://memgraph.com/docs/data-streams#start-a-stream) query that
  returns the raw message received by the transformation. [#392](https://github.com/memgraph/memgraph/pull/392)

#### Bug Fixes

- Fix header on `SHOW REPLICATION ROLE` query and wrong timeout info on `SHOW
  REPLICAS` query. [#376](https://github.com/memgraph/memgraph/pull/376)
- Fix WebSocket connection with clients that do not use binary protocol header. [#403](https://github.com/memgraph/memgraph/pull/403)
- Fix SSL connection shutdown hanging. [#395](https://github.com/memgraph/memgraph/pull/395)
- Fix module symbol loading with python modules. [#335](https://github.com/memgraph/memgraph/pull/335)
- Fix header on `SHOW REPLICATION ROLE` query and wrong timeout info on
  `SHOW REPLICAS query`. [#376](https://github.com/memgraph/memgraph/pull/376)
- Adapted compilation flag so that the memory allocator uses JEMALLOC while
  counting allocated memory. [#401](https://github.com/memgraph/memgraph/pull/401)

### Lab v2.1.1 - May 27, 2022

#### What's New

* Add tooltips and highlights throughout the application.

#### Bug Fixes

* Fix several bugs with streams.

### Lab v2.1.0 - May 25, 2022

#### What's New

* Add the ability to view, create, edit, start, stop, test, and remove streams.
* Add a new connecting screen with the ability to set monitoring (logs) port.
* Add Cypher query persistence when closing/opening Cypher query editor.
* Add node label, relationship type, and node/relationship property Cypher code suggestions for small graphs (number of nodes < 100k and number of relationships < 200k).
* Add module function Cypher code suggestions.
* Add module support for adding functions along with `mgp` suggestions and documentation.
* Add new GSS graph functions: `InEdges`, `OutEdges`, `Edges`, `AdjacentNodes`, `StartNode`, `EndNode`, `NodeCount`, `EdgeCount`.
* Add new GSS array functions: `RandomOf`, `Find`, `Filter`, `Map`, `All`, `Any`, `Uniq`.

#### Bug Fixes

* Fix the UI for the GSS error messages.
* Fix the Cypher code suggestion for modules with `.` in the namespace name.
* Fix several bugs with query collections.
* Fix the empty states across the application.
* Fix the import progress bar.
* Fix the graph schema for an empty database.
* Fix the responsiveness across the application.
* Add the maximum limit of five vertical layouts.
* Fix the loading issue when running multiple Cypher queries at once.

### MAGE v1.3.0 - May 23, 2022

#### Major Features and Improvements

- Added integration between cuGraph and Memgraph integration.
  [#99](https://github.com/memgraph/mage/pull/99)

#### Bug Fixes

- Fixed node deletion.
  [#141](https://github.com/memgraph/mage/pull/141)

### Memgraph v2.3.0 - Apr 27, 2022

#### Major Features and Improvements

- Added [`FOREACH`](https://memgraph.com/docs/querying/clauses) clause.
  [#351](https://github.com/memgraph/memgraph/pull/351)
- Added [Bolt over WebSocket](https://memgraph.com/docs/client-libraries/javascript) support to
  Memgraph. [#384](https://github.com/memgraph/memgraph/pull/384)
- Added [user-defined Memgraph magic
  functions](https://memgraph.com/docs/querying/functions#user-defined-memgraph-magic-functions).
  [#345](https://github.com/memgraph/memgraph/pull/345)

#### Bug Fixes

- Fixed incorrect loading of C query modules.
  [#387](https://github.com/memgraph/memgraph/pull/387)

### Lab v2.0.3 - Apr 27, 2022

#### Bug Fixes

* Fix the encrypted connection creation towards Memgraph.
* Fix duplicate keywords in Cypher and Python code suggestion tools.

### Lab v2.0.2 - Apr 22, 2022

#### Major Features and Improvements

- Add guides for empty states throughout the app.
- Add an ability to close hints for transformations and procedures in module view.
- Add an ability to download query results in JSON format.
- Add confirmation step for all delete actions throughout the app.
- Add the generic Cypher query as a sample query after custom dataset file import.

#### Bug Fixes

- Fix the table view with a better resize functionality throughout the app.
- Change the color of the node labels and relationship types in the Cypher query editor.
- Fix the delete query collection action.
- Fix opening an external link in the browser instead of the Lab app.
- Fix the initial render of the map for geo graph results.
- Replace the toast message "Web socket stopped working" with better notice in the "Logs" view.

### MAGE v1.2.0 - Apr 20, 2022

#### Major Features and Improvements

- Implemented Temporal graph networks.
  [#121](https://github.com/memgraph/mage/pull/121)
- Implemented Dynamic Betweenness Centrality.
  [#127](https://github.com/memgraph/mage/pull/127)
- Implemented Dynamic Katz Centrality.
  [#117](https://github.com/memgraph/mage/pull/117)
- Implemented Louvain Community Detection.
  [#48](https://github.com/memgraph/mage/pull/48)
- Implemented Maximum Flow.
  [#125](https://github.com/memgraph/mage/pull/125)
- Implemented Static Katz Centrality.
  [#117](https://github.com/memgraph/mage/pull/117)
- Added utility Import/Export module (JSON).
  [#100](https://github.com/memgraph/mage/pull/100)
- Bumped the version of Black formatter.
  [#132](https://github.com/memgraph/mage/pull/132)

#### Bug Fixes

- Fixed IsSubset checking for unordered set.
  [#135](https://github.com/memgraph/mage/pull/135)
- Fixed Continuous integration.
  [#133](https://github.com/memgraph/mage/pull/133)
- Fixed E2E testing.
  [#128](https://github.com/memgraph/mage/pull/128)
- Fixed ID validity check.
  [#129](https://github.com/memgraph/mage/pull/129)

### Lab v2.0.1 - Apr 8, 2022

#### Major Features and Improvements

- Add context (graph schema, description) to each dataset template.
- Add an action to download query results.

#### Bug Fixes

- Fix the bug when adding a query to the query collection.
- Fix several typos and copies.
- Fix the web socket connection issue for the manual Memgraph connect.
- Fix initial code suggestions which are dependent on the Memgraph version.

### Lab v2.0.0 - Mar 31, 2022

#### Major Features and Improvements

- Add horizontal and vertical layouts for custom layout configuration.
- Add more query information in the latest queries: runtime, status, number of
  results.
- Add query collections to structure and save favorite queries.
- Add better Cypher code suggestion for functions, modules, nodes,
  relationships, properties.
- Add Cypher code documentation on highlight.
- Add Graph Style Script code suggestion for `@NodeStyle`, `@EdgeStyle`,
  properties and functions.
- Add Graph Style Script code documentation on highlight.
- Add improved table views throughout the app.
- Add new rendering and simulation engine based on D3.js.
- Add new rendering simulation options: collision, repel force and link
  distance.
- Remove definition of query parameters when running a Cypher query with
  `$variable`.
- Add real-time logs view from Memgraph.
- Add a status tray with connection status and main Memgraph metrics.
- Add real-time connection status and automatic reconnect ability.
- Add new graph schema view with distribution of present properties in
  nodes/relationships.
- Add ability to view, edit, remove and change query modules.

### Memgraph v2.2.1 - Mar 17, 2022

#### Bug Fixes

- Added CentOS 7 release by fixing the compatibility issue with the older
  version of SSL used on CentOS 7.
  [#361](https://github.com/memgraph/memgraph/pull/361)

### Memgraph v2.2.0 - Feb 18, 2022

#### Major Features and Improvements

- Added support for compilation on ARM architectures (aarch64) and Docker
  support for running Memgraph on Apple M1 machines.
  [#340](https://github.com/memgraph/memgraph/pull/340)
- Added [monitoring server](https://memgraph.com/docs/configuration/monitoring-server) that forwards
  certain information from Memgraph to the clients connected to it (e.g. logs)
  using WebSocket. [#337](https://github.com/memgraph/memgraph/pull/337)
- Added `CONFIGS` and `CREDENTIALS` options to [Kafka streams](https://memgraph.com/docs/data-streams).
  [#328](https://github.com/memgraph/memgraph/pull/328)
- Added [built-in procedures used for handling Python module
  files](https://memgraph.com/docs/custom-query-modules/manage-query-modules).
  `mg.create_module_file`, `mg.update_module_file`, `mg.delete_module_file`,
  `mg.get_module_file`, and `mg.get_module_files` allow you to do modifications
  on your Python module files, get their content, and list all the files present
  in your query module directories directly from Memgraph.
  [#330](https://github.com/memgraph/memgraph/pull/330)
- Built-in procedures
  [`mg.procedures`](https://memgraph.com/docs/custom-query-modules/manage-query-modules#utility-query-module) and
  [`mg.transformations`](https://memgraph.com/docs/data-streams/transformation-modules#utility-procedures-for-transformations)
  return additional information about the procedures and transformations
  scripts. `path` returns an absolute path to the module file containing the
  procedure, while `is_editable` returns `true` if the file can be edited using
  Memgraph or `false` otherwise.
  [#310](https://github.com/memgraph/memgraph/pull/310)
- [Added `SHOW VERSION` query](https://memgraph.com/docs/configuration/server-stats) that returns the version of the Memgraph server
  which is being queried. [#265](https://github.com/memgraph/memgraph/pull/265)

#### Bug Fixes

- The reference count is increased when `Py_None` is returned from the `_mgp`
  module. This fixes a nondeterministic fatal Python error.
  [#320](https://github.com/memgraph/memgraph/pull/320)
- Use correct error when printing warning in rebalance callback of Kafka
  consumer. [#321](https://github.com/memgraph/memgraph/pull/321)
- Fix transaction handling in streams in case of serialization error.
  Previously, a serialization error caused an exception to be thrown since
  nested transactions are not supported. After this fix, the transactions are
  handled correctly in the transaction retry logic.
  [#339](https://github.com/memgraph/memgraph/pull/339)
- Temporal types `LocalTime` and `LocalDateTime` instantiations return subsecond
  precision. Additionally, query modules `mg_local_date_time_now()` and
  `mg_local_time_now()` also return subsecond precision.
  [#333](https://github.com/memgraph/memgraph/pull/333)

### MAGE v1.1.0 - Dec 13, 2021

#### Major Features and Improvements

- Updated rsmgp-sys to the new MGP API.
  [#78](https://github.com/memgraph/mage/pull/78)
- Add temporal type to rsmgp-sys.
  [#82](https://github.com/memgraph/mage/pull/82)
- Implemented node2vec. [#81](https://github.com/memgraph/mage/pull/81)
- Updated GraphView abstraction. [#85](https://github.com/memgraph/mage/pull/85)
- Implemented approximative streaming PageRank.
  [#69](https://github.com/memgraph/mage/pull/69)
- Implemented weighted graph methods built for dynamic community detection.
  [#89](https://github.com/memgraph/mage/pull/89)
- Implemented LabelRankT dynamic community detection algorithm.
  [#66](https://github.com/memgraph/mage/pull/66)

#### Bug Fixes

- Fixed memory leakage. [#77](https://github.com/memgraph/mage/pull/77)
- Solved dependency vulnerability.
  [#83](https://github.com/memgraph/mage/pull/83)
- Fixed `set_cover.greedy` result type bug.
  [#76](https://github.com/memgraph/mage/pull/76)
- Fixed MAGE installation on Linux based distro.
  [#92](https://github.com/memgraph/mage/pull/92)

### Memgraph v2.1.1 - Dec 07, 2021

#### Breaking Changes

- Loading streams created by versions of Memgraph older than 2.1 is not
  possible. We suggest you extract the necessary information using the older
  version of Memgraph and recreate the streams in a newer version (Memgraph 2.1
  and newer).

#### Major Features and Improvements

- Added procedures for retrieving configuration information specific for each
  stream type. `mg.pulsar_stream_info` will return information about a specific
  Pulsar stream and `mg.kafka_stream_info` will return information about a
  specific Kafka stream. [#301](https://github.com/memgraph/memgraph/pull/301)
- `SHOW STREAMS` now returns default values for batch interval and batch size if
  they weren't specified. [#306](https://github.com/memgraph/memgraph/pull/306)

#### Bug Fixes

- Query execution stats, returned after a Cypher query was executed, are now
  updated with the changes made in write procedures.
  [#304](https://github.com/memgraph/memgraph/pull/304)
- Loading streams created by older versions won't cause Memgraph to crash.
  [#302](https://github.com/memgraph/memgraph/pull/302)

### Lab v1.3.6 - Dec 3, 2021

#### Bug Fixes

* Fix the bug when returning edges: `Cannot read properties of undefined (reading 'push')`.

### Memgraph v2.1.0 - Nov 22, 2021

#### Breaking Changes

- Loading streams created by older versions cause Memgraph to crash. The only
  possible workaround involves **deleting the existing streams**.
  The streams can be deleted by the `DROP STREAM` query in the old versions of
  Memgraph. After upgrading to this version, the `streams` directory has to be
  deleted manually from Memgraph's data directory (on Debian-based systems, it
  is `/var/lib/memgraph` by default).
- The query for creating a Kafka stream now requires the `KAFKA` keyword. The
  previous form `CREATE STREAM ...` was changed to `CREATE KAFKA STREAM ...`.

#### Major Features and Improvements

- Now supporting Bolt protocol version 4.3.
  [#226](https://github.com/memgraph/memgraph/pull/226)
- Streams support for retrying conflicting transactions. When a message is
  processed from a certain stream source, a query is executed as a part of the
  transaction. If that transaction fails because of other conflicting
  transactions, the transaction is retried a set number of times. The number of
  retries and interval between each retry can be controlled with configs
  `--stream-transaction-conflict-retries` and
  `--stream-transaction-retry-interval`.
  [#294](https://github.com/memgraph/memgraph/pull/294)
- Added procedure to configure the starting offset (to consume messages from) of
  a topic (and its partitions).
  [#282](https://github.com/memgraph/memgraph/pull/282)
- Added `BOOTSTRAP_SERVERS` option to `CREATE KAFKA STREAM` which you can check
  [here](https://memgraph.com/docs/data-streams).
  [#282](https://github.com/memgraph/memgraph/pull/282)
- Added Bolt notifications in the query summary to inform the user about results
  or to give useful tips. When a query executes successfully, sometimes is
  necessary to give users tips or extra information about the execution.
  [#285](https://github.com/memgraph/memgraph/pull/285)
- Added execution statistics in the query summary to inform user on how many
  objects were affected. E.g., when you run a query with a `CREATE` clause,
  you'll know how many nodes/edges were created by it.
  [#285](https://github.com/memgraph/memgraph/pull/285)
- Added support for connecting to Pulsar as a new stream source. For more
  details, check out our [reference pages](https://memgraph.com/docs/data-streams).
  [#293](https://github.com/memgraph/memgraph/pull/293)

#### Bug Fixes

- Allow duration values to be used as weights in the [Weighted Shortest
  Path](https://memgraph.com/docs/advanced-algorithms/deep-path-traversal#weighted-shortest-path)
  query. [#278](https://github.com/memgraph/memgraph/pull/278)
- Fix linkage error when `mgp_local_time_get_minute` is used.
  [#273](https://github.com/memgraph/memgraph/pull/273)
- Fix crash when temporal types are used with `ORDER BY` clause.
  [#299](https://github.com/memgraph/memgraph/pull/299)

### Lab v1.3.5 - Nov 17, 2021

#### What's New

* Add new Cypher stream keywords from Memgraph 2.1.0 release.

#### Bug Fixes

* Fix the copy to the clipboard bug to keep new lines.

### Lab v1.3.4 - Nov 15, 2021

#### What's New

* Add quick connect for Memgraph running locally.
* Add guides on how to install Memgraph locally.

### Lab v1.3.3 - Oct 22, 2021

#### Bug Fixes

- Fixed the action of exporting the database to a `cypherl` file.
- Added support for the temporal types in query responses.

### Memgraph v2.0.1 - Oct 12, 2021

#### Major Features and Improvements

- Updated a startup message with a link to the [getting started
  page](https://memgraph.com/docs/getting-started).
  [#259](https://github.com/memgraph/memgraph/pull/259)
- Updated certain error and warning messages in the logs with links to the
  documentation explaining the problem in more detail.
  [#243](https://github.com/memgraph/memgraph/pull/243)
- Updated mgconsole to
  [v1.1.0](https://github.com/memgraph/mgconsole/releases/tag/v1.1.0).
  [#260](https://github.com/memgraph/memgraph/pull/260)

#### Bug Fixes

- Graph updates made in the write procedures are now correctly registered in the
  triggers. [#262](https://github.com/memgraph/memgraph/pull/262)
- Fixed `DETACH DELETE` interaction with the triggers. Previously, vertices
  deleted by the `DETACH DELETE` would not be registered by triggers if only `ON () DELETE` trigger existed.
  [#266](https://github.com/memgraph/memgraph/pull/266)

### Memgraph v2.0.0 - Oct 5, 2021

#### Breaking Changes

- Changed the `timestamp()` function to return `microseconds` instead of
  `milliseconds`.
- Most of the query modules C API functions are changed to return a `mgp_error`
  as a more fine-grained way of error reporting. The only exceptions are the
  functions that free allocated memory (`mgp_free` and `mgp_global_free`) and
  destroy objects (`mgp_value_destroy`, `mgp_list_destroy`, etc.) which
  remain the same.
- The first user created using the `CREATE USER` query will have all the
  privileges granted to him. Previously, you could've locked yourself out of
  Memgraph by creating a user and immediately disconnecting.

#### Major Features and Improvements

- Added support for temporal types, a feature that allows the user to manipulate
  and store time related data in the graph. For more information take a look at
  the [reference guide](https://memgraph.com/docs/fundamentals/data-types)
- Added support for parameters with `CREATE` clause in the following form:
  `CREATE (n $param)`.
- Introduced settings to Memgraph that can be modified during runtime. You can
  check out more details [here](https://memgraph.com/docs/configuration/enabling-memgraph-enterprise).
- Added writeable procedure support, so
  [procedures](https://memgraph.com/docs/custom-query-modules/python/python-example)
  can modify the graph by creating and deleting vertices and edges, modifying
  the labels of vertices or setting the properties of vertices and edges.

#### Bug Fixes

- Fixed planning of queries with `MERGE` clause. If a previously defined symbol
  is used as property value inside the `MERGE` clause, the planner will
  correctly use the label-property index if present.
- Unused memory is correctly returned to OS when `FREE MEMORY` query is used.
  Before, Memgraph would free up the memory internally and not return it to the
  OS. Because of that Memgraph could allocate more memory from the OS than it's
  allowed.
- Fixed recovery from durability files. Because of a wrong check, Memgraph could
  crash and leave the durability files in an invalid state making recovery not
  possible.
- Fixed usage of `execute` keyword in queries. Because of the special way we
  handle the `EXECUTE` keyword from the `CREATE TRIGGER` query using that same
  keyword in other contexts caused Memgraph to crash.

### Lab v1.3.2 - Oct 5, 2021

#### Bug Fixes

- Fixed the copy to clipboard bug with removed spaces.
- Updated the Cypher IntelliSense with the latest commands.

### Lab v1.3.1 - Sep 27, 2021

#### Major Features and Improvements

- Signed the Memgraph Lab applications for macOS and Windows.

#### Bug Fixes

- Fixed the paste overwrite action in the query editor.
- Fixed the bug `Cannot read property 'class' of null`.

### Memgraph v1.6.1 - Jul 24, 2021

#### Major Features and Improvements

- Added proper privilege checks for queries executed by triggers and stream
  transformations.

#### Bug Fixes

- Fixed error handling in streams to make restarting streams possible after
  failing. The issue is caused by not rolling back the transaction in which the
  query failed, so when the stream was restarted and tried to process the next
  batch of messages it was still in a transaction, but it tried to start a new
  one. Now the transaction is rolled back in case of any errors during query
  execution, so a new transaction can be started during the processing of the
  next batch of messages.

### Memgraph v1.6.0 - Jul 7, 2021

#### Breaking Changes

- Changed the `LOCK_PATH` permission to `DURABILITY`.

#### Major Features and Improvements

- Added support for consuming Kafka streams. You can connect Memgraph to a Kafka
  cluster and run queries based on the messages received. The transformation
  from Kafka to Cypher queries is done using **Transformation Modules**, a
  concept similar to Query Modules. Using our Python and C API, you can easily
  define functions that analyze Kafka messages and generate different queries
  based on them. The stream connection can be configured, tested, stopped,
  started, checked, and dropped.
- Introduced global allocators for Query Modules using C API, so the data can be
  preserved between multiple runs of the same procedure.
- Introduced new isolation levels, `READ COMMITTED` and `READ_UNCOMMITTED`. The
  isolation level can be set with a config. Also, you can set the isolation
  level for a certain session or the next transaction. The names of the
  isolation levels should be self-explanatory, unlike the `SNAPSHOT ISOLATION`
  which is still the default isolation level.
- The query timeouts are now triggered using a different method. Before, we used
  the TSC to measure the execution time. Unfortunately, this proved unreliable
  for certain CPUs (AMD Ryzen 7 and M1), which caused queries to timeout almost
  instantly. We switched to POSIX timer which **should** work on every hardware,
  while not affecting the performance.
- Added a config, `allow-load-csv`, with which you can disable `LOAD CSV`
  clause. `LOAD CSV` can read and display data from any file on the system which
  could be insecure for some systems. Because of that, we added a config that
  allows you to disable that clause in every case.
- Added `CREATE SNAPSHOT` query. Snapshots are created every few minutes, using
  this query you can trigger snapshot creation instantly.
- Increased the default query timeout to 10 minutes. The previous default amount
  of 3 minutes proved too small, especially for queries that use `LOAD CSV` with
  a large dataset.

#### Bug Fixes

- Fixed parsing of certain types in Query Modules using Python API.
- Fixed a concurrency bug for Query Modules using Python API. Running the same
  procedure from multiple clients caused the Memgraph instance to crash.
- Fixed restoring triggers that call procedures. Because the triggers were
  restored before the procedures, the query trigger executes couldn't find the
  called procedure, which caused the restore to fail. Switching up the order was
  enough to fix the problem.

### Memgraph v1.5.0 - May 28, 2021

#### Major Features and Improvements

- Added database triggers. You can now create, delete, and print out triggers
  that execute Cypher statements. You can create custom actions whenever a node
  or an edge is created, updated, or deleted. All the triggers are persisted on
  the disk, so no information is lost between runs.
- Replaced mg_client with the mgconsole command-line interface, which ships
  directly with Memgraph. You can now install mgconsole directly on Windows and
  macOS.

#### Bug Fixes

- Fixed parsing of types for Python procedures for types nested in `mgp. List`.
  For example, parsing of `mgp.List[mgp.Map]` works now.
- Fixed memory tracking issues. Some of the allocation and deallocation wasn't
  tracked during the query execution.
- Fixed reading CSV files that are using CRLF as the newline symbol.
- Fixed permission issues for `LOAD CSV`, `FREE MEMORY`, `LOCK DATA DIRECTORY`,
  and replication queries.

### Memgraph v1.4.0 - Apr 2, 2021

#### Breaking Changes

- Changed `MEMORY LIMIT num (KB|MB)` clause in the procedure calls to `PROCEDURE MEMORY LIMIT num (KB|MB)`. The functionality is still the same.

#### Major Features and Improvements

- Added replication to community version.
- Added support for multiple query modules directories at the same time. You can
  now define multiple, comma-separated paths to directories from which the
  modules will be loaded using the `--query-modules-directory` flag.
- Added support for programmatically reading in data from CSV files through the
  `LOAD CSV` clause. We support CSV files with and without a header, the
  supported dialect being Excel.
- Added a new flag `--memory-limit` which enables the user to set the maximum
  total amount of memory memgraph can allocate during its runtime.
- Added `FREE MEMORY` query which tries to free unused memory chunks in
  different parts of storage.
- Added the memory limit and amount of currently allocated bytes in the result
  of `SHOW STORAGE INFO` query.
- Added `QUERY MEMORY LIMIT num (KB|MB)` to Cypher queries which allows you to
  limit memory allocation for the entire query. It can be added only at the end
  of the entire Cypher query.
- Added logs for the different parts of the recovery process. `INFO`, `DEBUG`
  and `TRACE` level all contain additional information that is printed out while
  the recovery is in progress.

#### Bug Fixes

- Fixed garbage collector by correctly marking the oldest current timestamp
  after the database was recovered using the durability files.
- Fixed reloading of the modules with changed result names.
- Fixed profile query to show the correct name of the ScanAll operator variant.

### Lab v1.3.0 - Feb 19, 2021

#### Major Features and Improvements

- Added option to show predefined datasets with the ability to import them to
  Memgraph.
- Added option to show sample query for every loaded predefined dataset.
- Added import of custom Cypher file datasets (`cypherl` format).
- Added export of current database state to Cypher file (`cypherl` format).
- Added default node label in graph view if name property is missing.
- Added default relationship type label in graph view for smaller graphs.

#### Bug Fixes and Other Changes

- Fixed sidebar links in the browser Lab.
- Fixed columns in favorite queries view.
- Fixed showing large amounts of properties in a popup when viewing node details
  in the graph view.
- Fixed the label in the popup when switching between edges and nodes in the
  graph view.
- Fixed node count in the dashboard view.
- Added descriptive and better error messages when connecting to Memgraph with
  encryption on/off.
- Fixed the close button in a node popup in the graph view.
- Fixed the spacing of the close button and relationship type in a relationship
  popup in the graph view.
- Fixed storing physics and styles across multiple query runs.
- Fixed initial positioning in graph view when running query in the data view.
- Fixed graph view reset when a query on data view had no results to show.
- Fixed map disappearing when running query multiple times in a row.
- Fixed running multiple Lab instances of the application on Windows and Linux.
- Fixed node size and spacing in graph view when showing smaller graphs.
- Fixed transition state issues between graph view and data view.

### Memgraph v1.3.0 - Jan 26, 2021

#### Breaking Changes

- Added extra information in durability files to support replication, making it
  incompatible with the durability files generated by older versions of
  Memgraph. Even though the replication is an Enterprise feature, the files are
  compatible with the Community version.

#### Major Features and Improvements

- Added support for data replication across a cluster of Memgraph instances.
  Supported instance types are MAIN and REPLICA. Supported replication modes are
  SYNC (all SYNC REPLICAS have to receive data before the MAIN can commit the
  transaction), ASYNC (MAIN doesn't care if data is replicated), SYNC WITH
  TIMEOUT (MAIN will wait for REPLICAS within the given timeout period, after
  timeout, replication isn't aborted but the replication demotes the REPLICA to
  the ASYNC mode).
- Added support for query type deduction. Possible query types are `r` (read),
  `w` (write), `rw` (read-write). The query type is returned as a part of the
  summary.
- Improved logging capabilities by introducing granular logging levels. Added
  new flag, `--log-level`, which specifies the minimum log level that will be
  printed. E.g., it's possible to print incoming queries or Bolt server states.
- Added ability to lock the storage data directory by executing the `LOCK DATA DIRECTORY` query which delays the deletion of the files contained in the
  data directory. The data directory can be unlocked again by executing the
  `UNLOCK DATA DIRECTORY` query.

### Lab v1.2.0 - Nov 3, 2020

#### Major Features and Improvements

- Added ability to create custom graph styling for nodes and edges in graph view
  with graph style language (similar to CSS).
- Added ability to save and load custom graph styling.
- Added ability to show map background for nodes with lat and lng numeric
  properties.
- Added ability to change map background style.
- Removed edge labels to be shown by default in graph view.
- Fixed overall UI and UX.
- Set encrypted connection to be turned off by default on login screen (Memgraph
  v1.2.0 comes with SSL off by default).

#### Bug Fixes

- Added ability to hide graph view if there are no node/edge data in response.

### Memgraph v1.2.0 - Oct 20, 2020

#### Breaking Changes

- SSL is disabled by default (`--bolt-cert-file` and `--bolt-key-file are
  empty`). This change might only affect the client connection configuration.

#### Major Features and Improvements

- Added support for Bolt v4.0 and v4.1.
- Added `mgp_networkx.py` as an alternative implementation of NetworkX graph
  objects, which is useful to use Memgraph data from NetworkX algorithms
  optimally.
- Added `nxalg.py` query module as a proxy to NetworkX algorithms.
- Added plan optimization to use a label-property index where the property is
  not null. As a result, the query engine, instead of scanning all elements and
  applying the filter, performs a label-property index lookup when possible.

#### Bug Fixes and Other Changes

- Fixed Cypher `ID` function `Null` handling. When the `ID` function receives
  `Null`, it will also return `Null`.
- Fixed bug that caused random crashes in SSL communication on platforms that
  use older versions of OpenSSL (< 1.1) by adding proper multi-threading
  handling.
- Fix `DISCARD` message handling. The query is now executed before discarding
  the results.

### Memgraph v1.1.0 - Jul 1, 2020

#### Major Features and Improvements

- Properties in nodes and edges are now stored encoded and compressed. This
  change significantly reduces memory usage. Depending on the specific dataset,
  total memory usage can be reduced up to 50%.
- Added support for rescanning query modules. Previously, the query modules
  directory was scanned only upon startup. Now it is scanned each time the user
  requests to load a query module. The functions used to load the query modules
  were renamed to `mg.load()` and `mg.load_all()` (from `mg.reload()` and
  `mg.reload_all()`).
- Improved execution performance of queries that have an IN list filter by using
  label+property indexes. Example: `MATCH (n: Label) WHERE n.property IN [] ...`

- Added support for `ANY` and `NONE` openCypher functions. Previously, only
  `ALL` and `SINGLE` functions were implemented.

#### Bug Fixes and Other Changes

- Fixed invalid paths returned by variable expansion when the starting node and
  destination node used the same symbol. Example: `MATCH path = (n: Person {name: "John"})-[: KNOWS*]->(n) RETURN path`

- Improved semantics of `ALL` and `SINGLE` functions to be consistent with
  openCypher when handling lists with `Null` s.
- `SHOW CONSTRAINT INFO` now returns property names as a list for unique
  constraints.
- Escaped label/property/edgetype names in `DUMP DATABASE` to support names with
  spaces in them.
- Fixed handling of `DUMP DATABASE` queries in multi-command transactions (
  `BEGIN`, ..., `COMMIT`).
- Fixed handling of various query types in explicit transactions. For example,
  constraints were allowed to be created in multi-command transactions (`BEGIN`
  , ..., `COMMIT`) but that isn't a transactional operation and as such can't
  be allowed in multi-command transactions.
- Fixed integer overflow bugs in `COUNT`, `LIMIT` and `SKIP`.
- Fixed integer overflow bugs in weighted shortest path expansions.
- Fixed various other integer overflow bugs in query execution.
- Added Marvel Comic Universe tutorial.
- Added FootballTransfers tutorial.

### Lab v1.1.3 - Jun 5, 2020

#### Bug Fixes

* Disable hardware acceleration.

### Lab v1.1.2 - Apr 10, 2020

#### Bug Fixes

* Fix side menu documentation and support links.

### Memgraph v1.0.0 - Apr 3, 2020

#### Major Features and Improvements

- [Enterprise Ed.] Exposed authentication username/rolename regex as a flag (
  `--auth-user-or-role-name-regex`).
- [Enterprise Ed.] Improved auth module error handling and added support for
  relative paths.
- Added support for Python query modules. This release of Memgraph supports
  query modules written using the already existing C API and the new Python API.
- Added support for unique constraints. The unique constraint is created with a
  label and one or more properties.
- Implemented support for importing CSV files (`mg_import_csv`). The importer is
  compatible with the Neo4j batch CSV importer.
- Snapshot and write-ahead log format changed (backward compatible with v0.50).
- Vertices looked up by their openCypher ID (`MATCH (n) WHERE ID(n) = ...`) will
  now find the node in O(logn) instead of O(n).
- Improved planning of BFS expansion, a faster, specific approach is now favored
  instead of a ScanAll+Filter operation.
- Added syntax for limiting memory of `CALL`.
- Exposed server name that should be used for Bolt handshake as flag (
  `--bolt-server-name-for-init`).
- Added several more functions to the query module C API.
- Implemented a storage locking mechanism that prevents the user from
  concurrently starting two Memgraph instances with the same data directory.

#### Bug Fixes and Other Changes

- [Enterprise Ed.] Fixed a bug that crashed the database when granting
  privileges to a user.
- [Enterprise Ed.] Improved Louvain algorithm for community detection.
- Type of variable expansion is now printed in `EXPLAIN` (e.g. ExpandVariable,
  STShortestPath, BFSExpand, WeightedShortestPath).
- Correctly display `CALL` in `EXPLAIN` output.
- Correctly delimit arguments when printing the signature of a query module.
- Fixed a planning issue when `CALL` preceded filtering.
- Fixed spelling mistakes in the storage durability module.
- Fixed storage GC indexes/constraints subtle race condition.
- Reduced memory allocations in storage API and indexes.
- Memgraph version is now outputted to `stdout` when Memgraph is started.
- Improved RPM packaging.
- Reduced number of errors reported in production log when loading query
  modules.
- Removed `early access` wording from the Community Offering license.

### Lab v1.1.1 - Apr 3, 2020

#### Bug Fixes

* Fix bug showing integers in node properties as strings.

### Lab v1.1.0

#### Major Features and Improvements

- Enable explain and profile view.
- Memgraph v0.15.0 keywords support.

#### Bug Fixes and Other Changes

- Fix bug with a new line in parsing multi-command queries.
- On empty data for graph redirect to data view.

### Lab v1.0.0

#### Major Features and Improvements

- Added unsecure connection option.
- Improved UX of login screen.
- Added basic tutorial that shows on the initial run.
- Added text search of history and favorite queries.
- Added storage statistics on overview screen.
- Added debug view with query explain and profile capabilities.
- Added graph schema (metagraph) generator.
- Improved query data (table) view.

### Lab v0.1.2

#### Bug Fixes and Other Changes

- Fixed app icon on MacOS.
- Improved error handling on the initial connect screen. Handle availability and
  secure connection errors.

### Lab v0.1.1

#### Major Features and Improvements

- Added overview view.
- Added query view (Monaco editor).
- Added graph, data and table data views.
- Added JSON export.
- Added electron builder packages for MacOS and Debian.

### Memgraph v0.50.0 - Dec 11, 2019

#### Breaking Changes

- [Enterprise Ed.] Remove support for Kafka streams.
- Snapshot and write-ahead log format changed (not backward compatible).
- Removed support for unique constraints.
- Label indexes aren't created automatically, create them explicitly instead.
- Renamed several database flags. Please see the configuration file for a list
  of current flags.

#### Major Features and Improvements

- [Enterprise Ed.] Add support for auth module.
- [Enterprise Ed.] LDAP support migrated to auth module.
- Implemented new graph storage engine.
- Add support for disabling properties on edges.
- Add support for existence constraints.
- Add support for custom openCypher procedures using a C API.
- Support loading query modules implementing read-only procedures.
- Add `CALL <procedure> YIELD <result>` syntax for invoking loaded procedures.
- Add `CREATE INDEX ON :Label` for creating label indexes.
- Add `DROP INDEX ON :Label` for dropping label indexes.
- Add `DUMP DATABASE` clause to openCypher.
- Add functions for treating character strings as byte strings.

#### Bug Fixes and Other Changes

- Fix several memory management bugs.
- Reduce memory usage in query execution.
- Fix bug that crashes the database when `EXPLAIN` is used.

### Memgraph v0.15.0 - Jul 17, 2019

#### Breaking Changes

- Snapshot and write-ahead log format changed (not backward compatible).
- `indexInfo()` function replaced with `SHOW INDEX INFO` syntax.
- Removed support for unique index. Use unique constraints instead.
- `CREATE UNIQUE INDEX ON :label (property)` replaced with `CREATE CONSTRAINT ON (n:label) ASSERT n.property IS UNIQUE`.
- Changed semantics for `COUNTER` openCypher function.

#### Major Features and Improvements

- [Enterprise Ed.] Add new privilege, `STATS` for accessing storage info.
- [Enterprise Ed.] LDAP authentication and authorization support.
- [Enterprise Ed.] Add audit logging feature.
- Add multiple properties unique constraint which replace unique indexes.
- Add `SHOW STORAGE INFO` feature.
- Add `PROFILE` clause to openCypher.
- Add `CREATE CONSTRAINT` clause to openCypher.
- Add `DROP CONSTRAINT` clause to openCypher.
- Add `SHOW CONSTRAINT INFO` feature.
- Add `uniformSample` function to openCypher.
- Add regex matching to openCypher.

#### Bug Fixes and Other Changes

- Fix bug in query comment parsing.
- Fix bug in query symbol table.
- Fix OpenSSL memory leaks.
- Make authentication case insensitive.
- Remove `COALESCE` function.
- Add movie tutorial.
- Add backpacking tutorial.

### Memgraph v0.14.1 - Jan 22, 2019

#### Bug Fixes and Other Changes

- Fix bug in explicit transaction handling.
- Fix bug in edge filtering by edge type and destination.

### Memgraph v0.14.0 - Oct 30, 2018

#### Breaking Changes

- Write-ahead log format changed (not backward compatible).

#### Major Features and Improvements

- [Enterprise Ed.] Reduce memory usage in distributed usage.
- Add `DROP INDEX` feature.
- Improve SSL error messages.

#### Bug Fixes and Other Changes

- [Enterprise Ed.] Fix issues with reading and writing in a distributed query.
- Correctly handle an edge case with unique constraint checks.
- Fix a minor issue with `mg_import_csv`.
- Fix an issue with `EXPLAIN`.

### Memgraph v0.13.0 - Oct 18, 2018

#### Breaking Changes

- Write-ahead log format changed (not backward compatible).
- Snapshot format changed (not backward compatible).

#### Major Features and Improvements

- [Enterprise Ed.] Authentication and authorization support.
- [Enterprise Ed.] Kafka integration.
- [Enterprise Ed.] Support dynamic worker addition in distributed.
- Reduce memory usage and improve overall performance.
- Add `CREATE UNIQUE INDEX` clause to openCypher.
- Add `EXPLAIN` clause to openCypher.
- Add `inDegree` and `outDegree` functions to openCypher.
- Improve BFS performance when both endpoints are known.
- Add new `node-label`, `relationship-type` and `quote` options to
  `mg_import_csv` tool.
- Reduce memory usage of `mg_import_csv`.

#### Bug Fixes and Other Changes

- [Enterprise Ed.] Fix an edge case in distributed index creation.
- [Enterprise Ed.] Fix issues with Cartesian in distributed queries.
- Correctly handle large messages in Bolt protocol.
- Fix issues when handling explicitly started transactions in queries.
- Allow openCypher keywords to be used as variable names.
- Revise and make user visible error messages consistent.
- Improve aborting time consuming execution.

### Memgraph v0.12.0 - Jul 4, 2018

#### Breaking Changes

- Snapshot format changed (not backward compatible).

#### Major Features and Improvements

- Improved Id Cypher function.
- Added string functions to openCypher (`lTrim`, `left`, `rTrim`, `replace`,
  `reverse`, `right`, `split`, `substring`, `toLower`, `toUpper`, `trim`
  ).
- Added `timestamp` function to openCypher.
- Added support for dynamic property access with `[]` operator.

### Memgraph v0.11.0 - Jun 20, 2018

#### Major Features and Improvements

- [Enterprise Ed.] Improve Cartesian support in distributed queries.
- [Enterprise Ed.] Improve distributed execution of BFS.
- [Enterprise Ed.] Dynamic graph partitioner added.
- Static nodes/edges id generators exposed through the Id Cypher function.
- Properties on disk added.
- Telemetry added.
- SSL support added.
- `toString` function added.

#### Bug Fixes and Other Changes

- Document issues with Docker on OS X.
- Add BFS and Dijkstra's algorithm examples to documentation.

### Memgraph v0.10.0 - Apr 24, 2018

#### Breaking Changes

- Snapshot format changed (not backward compatible).

#### Major Features and Improvements

- [Enterprise Ed.] Distributed storage and execution.
- `reduce` and `single` functions added to openCypher.
- `wShortest` edge expansion added to openCypher.
- Support packaging RPM on CentOS 7.

#### Bug Fixes and Other Changes

- Report an error if updating a deleted element.
- Log an error if reading info on available memory fails.
- Fix a bug when `MATCH` would stop matching if a result was empty, but later
  results still contain data to be matched. The simplest case of this was the
  query: `UNWIND [1, 2, 3] AS x MATCH (n: Label {prop: x}) RETURN n`. If there
  was no node `(: Label {prop: 1})`, then the `MATCH` wouldn't even try to find
  for `x` being 2 or 3.
- Report an error if trying to compare a property value with something that
  cannot be stored in a property.
- Fix crashes in some obscure cases.
- Commit log automatically garbage collected.
- Add minor performance improvements.

### Memgraph v0.9.0 - Dec 18, 2017

#### Breaking Changes

- Snapshot format changed (not backward compatible).
- Snapshot configuration flags changed, general durability flags added.

#### Major Features and Improvements

- Write-ahead log added.
- `nodes` and `relationships` functions added.
- `UNION` and `UNION ALL` is implemented.
- Concurrent index creation is now enabled.

#### Bug Fixes and Other Changes

### Memgraph v0.8.0

#### Major Features and Improvements

- CASE construct (without aggregations).
- Named path support added.
- Maps can now be stored as node/edge properties.
- Map indexing supported.
- `rand` function added.
- `assert` function added.
- `counter` and `counterSet` functions added.
- `indexInfo` function added.
- `collect` aggregation now supports Map collection.
- Changed the BFS syntax.

#### Bug Fixes and Other Changes

- Use \u to specify 4 digit codepoint and \U for 8 digit
- Keywords appearing in header (named expressions) keep original case.
- Our Bolt protocol implementation is now completely compatible with the
  protocol version 1 specification. (https://boltprotocol.org/v1/)
- Added a log warning when running out of memory and the
  `memory_warning_threshold` flag
- Edges are no longer additionally filtered after expansion.

### Memgraph v0.7.0

#### Major Features and Improvements

- Variable length path `MATCH`.
- Explicitly started transactions (multi-query transactions).
- Map literal.
- Query parameters (except for parameters in place of property maps).
- `all` function in openCypher.
- `degree` function in openCypher.
- User specified transaction execution timeout.

#### Bug Fixes and Other Changes

- Concurrent `BUILD INDEX` deadlock now returns an error to the client.
- A `MATCH` preceded by `OPTIONAL MATCH` expansion inconsistencies.
- High concurrency Antlr parsing bug.
- Indexing improvements.
- Query stripping and caching speedups.

### Memgraph v0.6.0

#### Major Features and Improvements

- AST caching.
- Label + property index support.
- Different logging setup & format.

### Memgraph v0.5.0

#### Major Features and Improvements

- Use label indexes to speed up querying.
- Generate multiple query plans and use the cost estimator to select the best.
- Snapshots & Recovery.
- Abandon old yaml configuration and migrate to gflags.
- Query stripping & AST caching support.

#### Bug Fixes and Other Changes

- Fixed race condition in MVCC. Hints exp+aborted race condition prevented.
- Fixed conceptual bug in MVCC GC. Evaluate old records w.r.t. the oldest.
  transaction's id AND snapshot.
- User friendly error messages thrown from the query engine.

## Build 837

#### Bug Fixes and Other Changes

- List indexing supported with preceeding IN (for example in query `RETURN 1 IN [[1, 2]][0]`).

## Build 825

#### Major Features and Improvements

- RETURN _, count(_), OPTIONAL MATCH, UNWIND, DISTINCT (except DISTINCT in
  aggregate functions), list indexing and slicing, escaped labels, IN LIST
  operator, range function.

#### Bug Fixes and Other Changes

- TCP_NODELAY -> import should be faster.
- Clear hint bits.

## Build 783

#### Major Features and Improvements

- SKIP, LIMIT, ORDER BY.
- Math functions.
- Initial support for MERGE clause.

#### Bug Fixes and Other Changes

- Unhandled Lock Timeout Exception.