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

Release 2.26.0

What's Changed

• Find endpoint returns wrong resource ids by @tarekziade in #1376
• Upgrade fastapi requirements to <=0.95.2 by @jotare in #1364
• Upgrade contributor app

Full Changelog: v2.25.0...v2.26.0

Release 2.25.0

What's Changed

• /chat ignores empty chunks at the end by @jotare in #1280
• properly utilize vector metadata to pull compose results by @vangheem in #1285
• do not preprocess query by @vangheem in #1286
• fix chat to always return relations by @vangheem in #1288
• Record metric on message sent to nats by @vangheem in #1289
• Include generated context in chat audit by @vangheem in #1290
• improved 1mstats and unify bench runner by @tarekziade in #1292
• add pod disruption budgets for components by @vangheem in #1291
• queue implementation missing task_done call by @vangheem in #1287
• fix standalone gc by @vangheem in #1300
• Deduplicate vector results by default in chat by @lferran in #1299
• catch and ignore errors in dir size by @tarekziade in #1301
• thread dump service rust by @hermeGarcia in #1270
• Parse status code from rephrase predict endpoint by @lferran in #1303
• Fix tokio task metrics counter increment bug by @jotare in #1306
• Use a subregistry to avoid prefix repetition by @jotare in #1307
• Fix num_remote_schedules metric name by @jotare in #1308
• Introduced a node_stack binary by @tarekziade in #1309
• set number of threads in paragraph index by @vangheem in #1310
• do not use rayon thread for search requests, fix deadlock by @vangheem in #1312
• Improve tokio metrics by @jotare in #1314
• moved to rust-spy by @tarekziade in #1322
• fix hpa version by @vangheem in #1327
• Add derived tokio metrics by @jotare in #1323
• be able to configure number of rayon threads by @vangheem in #1317
• use crossbeam threads by @vangheem in #1326
• fix deadlock from search implementation by @vangheem in #1321
• Proc macros crate and metrics measure macro by @jotare in #1250
• Rename duplicated metric name by @jotare in #1328
• Migrate node metrics to use new proc macro by @jotare in #1330
• be able to configure max_ack_pending by @vangheem in #1333
• Support synchronous chat response by @vangheem in #1331
• Export / Import logic by @lferran in #1329
• Integrate Export / Import into HTTP api by @lferran in #1335
• add search id to search request for logs by @vangheem in #1334
• nats msg progress updater by @vangheem in #1336
• nucliadb_texts integration tests by @jotare in #1339
• Bench runner against real shards by @tarekziade in #1324
• Enable processing in on prem by @bloodbare in #1294
• support pydantic 2 in telemetry by @vangheem in #1348
• Use measure macro in nucliadb_node by @jotare in #1350
• Ease install_requires for nucliadb_proto by @lferran in #1354
• Build test node Docker images in debug mode by @jotare in #1352
• fix node sidecar deps by @vangheem in #1359
• fixed nats leak in tests by @tarekziade in #1360
• Fixes the missing metadata on patch call by @tarekziade in #1361
• Improve list of stop words by @tarekziade in #1349
• add vm e2e test by @vangheem in #1358
• Local storage driver fixes by @lferran in #1363
• Run nucliadb tests on all maindb drivers by @lferran in #1346
• Upgrade Fast API by @bloodbare in #1368
• FST integration into DataPoint by @tarekziade in #1247
• Support more VM install tests through github actions by @vangheem in #1370
• cleanup measured paragraph cache implementation by @vangheem in #1373
• remove label from make file by @tarekziade in #1375
• added some debug calls by @tarekziade in #1377

Full Changelog: v2.24.0...v2.25.0

Release 2.24.0

open up sdk python support (#1279)

* open up sdk python support

* all libraries can be 3.8+

* python 3.11 in doc

* not tele for now

Release 2.23.0

What's Changed

• Clean Makefile by @vangheem in #1240
• Use SentryInterceptor in gRPC servers by @jotare in #1238
• Add SentryInterceptor to nucliadb_utils grpc server (II) by @jotare in #1242
• add lock file for installing nucliadb by @vangheem in #1241
• Explicit update policy for dependencies by @tarekziade in #1239
• Remove unused Rust dependencies by @jotare in #1243
• Fix predict chat response iterator by @lferran in #1231
• Upgrade to Python 3.11 by @jotare in #1246
• update contributing pytest notes by @vangheem in #1248
• nak errored messages by @vangheem in #1251
• Delete deprecated POST /entitiesgroups/{group} by @jotare in #1110
• Add the option to remove paragraph search on chat endpoint by @bloodbare in #1254
• Handle shard not found when getting shard info by @vangheem in #1257
• add middleware tracing by @hermeGarcia in #1258
• Remove_paragraph_from_chat_option by @bloodbare in #1259
• Fix wrong lazy loading logic by @jotare in #1261
• add logs to middleware by @hermeGarcia in #1263

Full Changelog: v2.22.0...v2.23.0

Release 2.22.0

What's Changed

• Better garbage collection scheduling by @vangheem in #1189
• implement count for kv driver by @vangheem in #1199
• convert number_of_shards into an async method by @tarekziade in #1192
• Prevent icon from being modified once is set by @lferran in #1201
• More chitchat/cluster-manager/cluster-monitor/swim cleanup by @lferran in #1203
• Promote extracted languages to basic.metadata by @lferran in #1197
• Cache implementation refactor by @vangheem in #1200
• Safer handling non existing keys by @lferran in #1204
• dont rely on the fs for vectorset updates by @hermeGarcia in #1205
• wait for the lock on GC by @hermeGarcia in #1210
• added RSS+VMS and disk size by @tarekziade in #1209
• Improve logging and add metric for garbage collecting by @vangheem in #1211
• Purge small fixes by @lferran in #1215
• make sure the reader sends telemetry by @tarekziade in #1202
• adjust default max paragraph and fields by @vangheem in #1214
• Retry timeout errors for tikv by @lferran in #1217
• on small kbs, calculate resources counter by @vangheem in #1220
• have ListShards list all the shards in a node by @vangheem in #1227
• Fix lock gc by @hermeGarcia in #1226
• Look ahead chat stream and parse predict status code by @lferran in #1228
• FST-based labels (initial class) by @tarekziade in #1225
• Add missing nonblocking code to GetShard by @jotare in #1233
• Avoid spawning unneeded threads on search by @jotare in #1235
• Standalone support for gc and shard splitting by @vangheem in #1222

Full Changelog: v2.21.0...v2.22.0