Release 2.37.0

What's Changed

• handle missing shard id by @vangheem in #1654
• Sidecar concurrent indexing by @jotare in #1570
• Fix gcloud overlapping log key by @jotare in #1656
• disable generation id cache by @vangheem in #1655
• fix generated version by @vangheem in #1659
• Added full coverage for async client by @lferran in #1651
• Signal dispatch captures errors instead of raising by @jotare in #1658
• auto versioning fix for rust by @vangheem in #1662
• Allow patching the resource slug by @lferran in #1657
• Trigger perf tests on merges to main by @lferran in #1653
• fix metadata usage to always use cache by @vangheem in #1661
• Garbage collection returns a status by @jotare in #1664
• Fix GC return value by @jotare in #1665
• Safer resource slug patching by @lferran in #1663
• Onprem fixes by @lferran in #1667
• add pg storage tests by @vangheem in #1669
• Fix audit chat when citations are present by @lferran in #1670
• implement read only txn support by @vangheem in #1660
• Return whether paragaph search results are fuzzy by @lferran in #1673
• safe delete in node by @hermeGarcia in #1668
• Small fixes and comments to relations2 by @jotare in #1675
• Allow users to pass generic context to LLM by @lferran in #1678
• Use new type annotation syntax by @jotare in #1677
• Fix type annotation by @lferran in #1680
• Removed bottleneck on search [standalone + pg] by @lferran in #1681
• Organize gitignore by @jotare in #1682
• Add index size property to counters response by @lferran in #1683
• Run ingest and search tests with s3 storage backend by @lferran in #1684
• add lead+ as code owner by @vangheem in #1666
• Read only tikv by @vangheem in #1674
• More resilient node fixture cleanup by @lferran in #1686
• Improve access log format and print absolute paths at startup by @lferran in #1685
• Add sorted paragraphs in find results by @lferran in #1687
• upgrade admin assets by @vangheem in #1689
• Change auth for bump GA by @rastut in #1693

Full Changelog: v2.36.0...v2.37.0

Release 2.36.0

decouple metadata from shard cache (#1635)

* decouple and cache shard generation ids

* move to metadata for management

* tweak

* fix test

* temp file when writing metadata

* fix

* try

* fix loading

* oops

Release 2.35.0

v2.35.0

bump version (#1644)

Release 2.34.0

fix conditional tikv dependency usage on telemetry (#1613)

* fix conditional tikv dependency usage on telemetry

* fix dep

Release 2.33.0

better nucliadb-telemetry requirements management (#1594)

* better nucliadb-telemetry requirements management

* try again

* moar

* fix

* fix

Release 2.32.0

Bump otel packages to resolve dependabot security alerts (#1584)

* Bump to resolve dependabot alert

* Bump to resolve dependabot alert

* Bump

* freeze reqs

Release 2.31.0

What's Changed

• Store and retrieve Q/A data by @tarekziade in #1478
• Cleanup rust settings management by @vangheem in #1544
• make setting config a patch by @vangheem in #1559
• Integrate cluster read replicas by @vangheem in #1535
• date range in vectors by @hermeGarcia in #1552
• support better filtering naming and origin metadata support by @vangheem in #1548
• Add user prompt to chat endpoint by @lferran in #1553
• Fix default parameters on search by @bloodbare in #1567
• Fix date range filtering by @lferran in #1566
• Improve progress reporting on imports by @lferran in #1574
• Fix file upload bug with pg storage driver by @lferran in #1575
• only set error log level when we are use file logging by @vangheem in #1577
• rename contributor into nucliadb admin by @ebrehault in #1576

Full Changelog: v2.30.0...v2.31.0

Release 2.30.0

What's Changed

• promote origin dates to indexed dates by @vangheem in #1518
• enable metrics on node read replicas by @vangheem in #1517
• support concurrent shard replication by @vangheem in #1489
• better replication healthy check by @vangheem in #1522
• rename usage of nucliadb analytics by @vangheem in #1523
• date range query by @hermeGarcia in #1526
• add reindex and reprocess to sdk by @vangheem in #1524
• better date range tests by @vangheem in #1530
• Create new training to stream ordered paragraphs by @jotare in #1531
• Be able to create a KB in experimental channel in stage by @lferran in #1533
• Add paragraph streaming to dataset tasks by @jotare in #1534
• Unify release channel in config and shards obj by @lferran in #1537
• After spec upload, trigger merge+deploy by @ebrehault in #1541
• graceful grpc server shutdown by @vangheem in #1532
• add read replica label by @vangheem in #1542
• fix merge trigger by @ebrehault in #1543
• Query planner and pre-filtering stage by @hermeGarcia in #1538
• Fix mypy on LRU type hints by @lferran in #1545
• Remove export import as task feature flag by @lferran in #1546
• Cut new version release by @lferran in #1547

Full Changelog: v2.29.0...v2.30.0

Release 2.29.0

What's Changed

• Add id to field and paragraph classifications by @jotare in #148
• better log handling support for standalone by @vangheem in #1482
• Resource cache and node timeout metrics by @lferran in #1491
• read replica k8s config by @vangheem in #1473
• Delete removed shards during replication by @vangheem in #1486
• support py3.8 on telemetry by @lferran in #1492
• adding perf to the docker image by @tarekziade in #1495
• sdk update_by_slug method by @clayrisse in #1499
• Fix documentation for resource slug endpoints by @tarekziade in #1445
• remove debug config for index node and just use RUST_LOG env var by @vangheem in #1507
• Forked Tantivy 0.17 by @hermeGarcia in #1510
• revert gc integration on sidecar(#1424) by @vangheem in #1513
• @clayrisse made their first contribution in #1499

Full Changelog: v2.28.0...v2.29.0

Release 2.27.0

What's Changed

• Export/Import v2 api by @lferran in #1385
• fix generating openapi spec by @vangheem in #1387
• [sdk] Be able to properly stream downloads to a file by @lferran in #1388
• Handle TiKV restarts by @vangheem in #1384
• Tasks api by @lferran in #1367
• Add deprecation warnings to nucliadb_client by @lferran in #1395
• Fix exporting conversations by @lferran in #1400
• Stdout logs improvements by @tarekziade in #1392
• release node bindings on every node change by @vangheem in #1396
• hnsw ops doc by @hermeGarcia in #1402
• update architecture diagram by @vangheem in #1397
• Remove nucliadb_client internal usages by @lferran in #1401
• handle none value in field serialize by @vangheem in #1404
• enable forking standalone process by @vangheem in #1282
• handle 429 on processing by @vangheem in #1405
• improve node pull logs by @vangheem in #1409
• try elock for gc by @hermeGarcia in #1410
• upgrade rust toolchaing by @hermeGarcia in #1411
• use single threaded searcher by @vangheem in #1398
• handle large nats messages by @vangheem in #1399

Full Changelog: v2.26.0...v2.27.0