| 8aa461b | David Turner | 19 July 2023, 13:24:09 UTC | Add docs on troubleshooting NFS repos (#97601) (#97812) Spell out a bit more clearly that ES works through the OS's filesystem abstraction, giving advice about how to reproduce problems outside of ES. | 19 July 2023, 13:24:09 UTC |
| c4d7657 | Ioana Tagirta | 19 July 2023, 12:43:07 UTC | Document dictionary parameter for Search Applications (#97501) (#97810) * Document dictionary parameter for Search Applications * Address PR feedback | 19 July 2023, 12:43:07 UTC |
| 9e281d6 | István Zoltán Szabó | 19 July 2023, 12:20:50 UTC | [DOCS] Adds important admonition to handling delayed data page (#97753) (#97805) | 19 July 2023, 12:20:50 UTC |
| 7136a06 | David Turner | 19 July 2023, 10:29:04 UTC | Fix NPE in Desired Balance API (#97775) (#97802) If a shard has no desired assignment then we cannot send the desired balance over the wire or render it as JSON because the `ShardAssignmentView` is `null`. This commit replaces the spurious `null` with an empty object, and cleans up the tests a little to remove some unnecessary mocking. | 19 July 2023, 10:29:04 UTC |
| 42c8402 | Iraklis Psaroudakis | 19 July 2023, 10:14:35 UTC | Mute testScheduleNowWithSystemClock (#97788) (#97796) Relates #95445 | 19 July 2023, 10:14:35 UTC |
| 0bdf0cf | Iraklis Psaroudakis | 19 July 2023, 10:06:45 UTC | [8.9] Mute testTimeThrottle (#97786) (#97791) * Mute testTimeThrottle (#97786) Relates #97518 * Update TimeThrottleIntegrationTests.java * Update TimeThrottleIntegrationTests.java | 19 July 2023, 10:06:45 UTC |
| 21b9bfb | István Zoltán Szabó | 19 July 2023, 09:02:11 UTC | [DOCS] Resoves conflict. (#97784) | 19 July 2023, 09:02:11 UTC |
| 63b8b9a | Iraklis Psaroudakis | 19 July 2023, 08:59:50 UTC | Mute some PkiAuthDelegationIntegTests (#97774) (#97777) Relates #97772 | 19 July 2023, 08:59:50 UTC |
| 361ac87 | Jack Conradson | 18 July 2023, 22:31:00 UTC | Fix `sub_searches` serialization bug (#97587) (#97767) We have allowed hybrid search since 8.4. This means the internal changes for sub_searches must be able to write a compound query as early as 8.4, but currently we only do that back to 8.8. This change fixes that issue. Closes ##97144 | 18 July 2023, 22:31:00 UTC |
| ff03c34 | Martijn van Groningen | 18 July 2023, 19:33:41 UTC | Change downsample threadpool size. (#97714) (#97764) Changed the downsample threadpool size to one eight of the allocated processors. Downsampling is a cpu intensive operation (search + heavy aggregation that rolls up documents) and could overwhelm a node. On top of this the downsample operation also delegates to force merge api, which has a threadpool that is also bounded to one eight of the allocated processors. Relates to #97141 | 18 July 2023, 19:33:41 UTC |
| 9728d3d | Ryan Ernst | 18 July 2023, 19:30:35 UTC | Remove transport version from main endpoint response (#97754) This is a partial backport of #97675. The transport version is a low level detail of how a node behaves in a cluster. It was recently added to the main endpoint response, but is too low level and should be moved to another endpoint TBD. This commit removes it from the main response, though keeps it as part of the wire format for the MainResponse to avoid any need for a bwc change to the format. | 18 July 2023, 19:30:35 UTC |
| 250307b | Mark Vieira | 18 July 2023, 18:11:16 UTC | Upgrade bundle JDK to Java 20.0.2 (#97758) (#97760) Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> | 18 July 2023, 18:11:16 UTC |
| 627ed6b | Ryan Ernst | 18 July 2023, 15:19:01 UTC | Muting pki tests until expired test certs are regenerated (#97757) see https://github.com/elastic/elasticsearch/issues/97756 | 18 July 2023, 17:24:18 UTC |
| 5b22118 | Daniel Mitterdorfer | 18 July 2023, 06:06:53 UTC | [Profiling] Fix off-by-one error in test assert (#97507) (#97511) The test case `testRandomLengthListGreaterThanSliceCount` is supposed to test behavior of a list slice implementation when the input list contains more elements than desired slices. However, there was an off-by-one error so sometimes the input list was exactly the same size as the desired number of slices and thus the assertion was wrong. With this commit we ensure that the input list is always larger than the desired number of slices. Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> | 18 July 2023, 06:06:53 UTC |
| 5774ac4 | Michael Peterson | 17 July 2023, 12:39:51 UTC | Allow users to run the painless execute API on a remote cluster shard (#97335) (#97694) Added a clusterAlias to the Painless execute Request object, so that index expressions in the request of the form "myremote:myindex" will be parsed to set clusterAlias to "myremote" and the index to "myindex". If clusterAlias is null, then it is executed against a shard on the local cluster, as before. If clusterAlias is non-null, then the SingleShardTransportAction is sent to the remote cluster, where it will run the full request (doing remote coordination). Note that the new clusterAlias field is not Writeable so that when it is sent to the remote cluster it will only see the index name, not the clusterAlias (which it wouldn't know how to handle correctly). Added PainlessExecuteIT test that tests cross-cluster calls Updated painless-execute-script end user docs to indicate support for cross-cluster executions Backport commit from cherry-pick of 6dd1841dbc2c754a7e9037dad6e42eba8f40c0cf from main (8.10-SNAPSHOT). Original PR: #97335 | 17 July 2023, 12:39:51 UTC |
| 3ce119d | Martijn van Groningen | 14 July 2023, 15:54:52 UTC | Document ILM's implicit rollover. (#97685) (#97696) This was introduced via #94065 Relates to #87246 | 14 July 2023, 15:54:52 UTC |
| c06ff05 | Craig Taverner | 14 July 2023, 12:57:38 UTC | Mute test for #97574 (#97652) Since it has failed repeatedly | 14 July 2023, 12:57:38 UTC |
| c332549 | Ignacio Vera | 14 July 2023, 09:23:55 UTC | adjust t-digest compression for MedianAbsoluteDeviationIT (#97680) (#97681) | 14 July 2023, 09:23:55 UTC |
| 391471d | Ignacio Vera | 14 July 2023, 01:17:24 UTC | Support for Byte and Short as vector tiles value tag (#97619) (#97672) Byte and short data types are not supported by the vector tiles specification. If a user tries to add one of those, it actually gets ignored. This add s support to those values by casting them as integers. In addition it will throw an exception if the value added is a type it cannot understand so it is not silently ignored. | 14 July 2023, 01:17:24 UTC |
| f501ef1 | Mayya Sharipova | 13 July 2023, 18:26:35 UTC | Add synonyms feature for test clusters (#97658) (#97666) Fix for #97334 where incorrect feature name was provided. Correct more instances of synonyms_feature_flag_enabled for synonyms_api_feature_flag_enabled Closes #96641, #97177 | 13 July 2023, 18:26:35 UTC |
| 9a886cc | Martijn van Groningen | 13 July 2023, 15:33:44 UTC | Explain tsdb counters better. (#97618) (#97662) | 13 July 2023, 15:33:44 UTC |
| 7f24fe3 | Andrei Stefan | 13 July 2023, 14:24:12 UTC | EQL: Softly remove "missing events" feature (#97643) (#97656) | 13 July 2023, 14:24:12 UTC |
| 36761f4 | Abdon Pijpelink | 13 July 2023, 12:49:49 UTC | Update filtering.asciidoc (#97542) (#97650) (cherry picked from commit 8f2760628fbda3c85d26adce1755f6b15506de27) Co-authored-by: Jean-Fabrice Bobo <2949987+jeanfabrice@users.noreply.github.com> | 13 July 2023, 12:49:49 UTC |
| ea6b823 | Mary Gouseti | 13 July 2023, 12:40:44 UTC | Test fix: ensure watcher history is green before searching (#97647) | 13 July 2023, 12:40:44 UTC |
| af2a1e8 | Abdon Pijpelink | 13 July 2023, 12:22:28 UTC | [DOCS] Clarify that dense vectors can be created with ES (#97636) (#97648) * [DOCS] Clarify that dense vectors can be created with ES * Fix rendering issue * Break up long sentence | 13 July 2023, 12:22:28 UTC |
| cce819d | Abdon Pijpelink | 13 July 2023, 08:30:51 UTC | [DOCS] Clarify that disk impact of TSDS varies per data set (#97571) (#97634) | 13 July 2023, 08:30:51 UTC |
| 35f3212 | István Zoltán Szabó | 12 July 2023, 07:56:16 UTC | [DOCS] Adds deployment_id as an option to query_vector_builder (#97576) (#97597) | 12 July 2023, 07:56:16 UTC |
| a4bab26 | David Roberts | 12 July 2023, 07:52:39 UTC | [ML] `inference_config` is optional for the infer trained model API (#97464) (#97598) It was made optional in #92359 which was released in version 8.6.1, but the docs weren't updated to reflect this. | 12 July 2023, 07:52:39 UTC |
| 3bed047 | Yang Wang | 12 July 2023, 01:12:11 UTC | Fix setting names in JWT realm doc. (#97545) (#97593) The claim patterns settings should be claim_patterns.xxx instead of claims.xxx_pattern. | 12 July 2023, 01:12:11 UTC |
| 0a00bc1 | Jack Conradson | 11 July 2023, 16:52:40 UTC | Mark rank and sub_searches as tech preview (#97573) (#97585) rank and sub_searches are in tech preview. This adds the tech preview text that is required in the docs for these features. | 11 July 2023, 16:52:40 UTC |
| c81dd7f | Chris Hegarty | 11 July 2023, 15:48:06 UTC | Update to Mockito 5.4.0 (#97285) (#97579) Backport of #97285 | 11 July 2023, 15:48:06 UTC |
| 726749f | István Zoltán Szabó | 11 July 2023, 10:25:08 UTC | [8.9][DOCS] Fixes typo in ELSER tutorial. (#97563) | 11 July 2023, 10:25:08 UTC |
| 66cf646 | Ievgen Degtiarenko | 11 July 2023, 08:26:54 UTC | Add `shard_stats.total_count` column description to /_cat/nodes docs. (#97549) (#97554) This column was implemented previously but not added to the docs. It might be helpful to list it when troubleshooting cluster balance. | 11 July 2023, 08:26:54 UTC |
| e2b5979 | David Turner | 11 July 2023, 07:19:18 UTC | Fix docs for explaining unassigned shards (#97538) (#97547) Today the `current_node` parameter is given in several sample requests illustrating how to explain an unassigned shard using the cluster allocation explain API. This doesn't make sense, an unassigned shard has no `current_node`. This commit removes the misleading parameter in these cases. | 11 July 2023, 07:19:18 UTC |
| 2c5f878 | Ignacio Vera | 11 July 2023, 00:36:55 UTC | Fix bug when creating empty geo_lines (#97509) (#97543) During reduce, only merge as non-overlapping if all geo-lines are non-overlapping. --------- Co-authored-by: Craig Taverner <craig@amanzi.com> | 11 July 2023, 00:36:55 UTC |
| c35543e | Alan Woodward | 10 July 2023, 13:55:48 UTC | Fix mixed-cluster test failure for top_hits aggregation (#97515) (#97521) The bug that this test checks is fixed is present in 8.8, so a mixed-cluster test using 8.8 will fail. | 10 July 2023, 13:55:48 UTC |
| c14fdf6 | Alan Woodward | 10 July 2023, 11:44:55 UTC | Set new providers before building FetchSubPhaseProcessors (#97460) (#97500) The FetchPhase sets source and field providers on its SearchLookup instance to ensure that we only have to load stored fields once per document. However, these were being set after subprocessors were constructed, and the FetchFieldsPhase subprocessor pulls a reference to the SearchLookup at construction time. This meant that the FieldsFetcher was not using these providers, instead using the standard source and fields lookup, in effect meaning that fields access would load source and fields separately. This would mostly just be an issue with performance; however, the top hits aggregation calls the fetch phase multiple times, with the result that the second and subsequent buckets in a top hits aggregation would see the source lookup from the previous bucket when building their FetchFieldsPhase subprocessor. This commit moves the call to SearchExecutionContext#setLookupProviders() in FetchPhase#buildSearchHits() before the call to build fetch phase subprocessors, ensuring that the fetch fields phase sees the correct SearchLookup. Fixes #96284 | 10 July 2023, 11:44:55 UTC |
| 3414aa7 | David Turner | 10 July 2023, 11:09:04 UTC | Add note on troubleshooting laggy cancellations (#97485) (#97498) Today we document that tasks may not react to cancellations immediately, but in practice it's surprising to users and kind of a bug if they run for too long after being cancelled. This commit adds a little extra detail about the information to collect to troubleshoot such a situation. | 10 July 2023, 11:09:04 UTC |
| 49c8327 | David Turner | 10 July 2023, 10:14:55 UTC | Enhance docs around network troubleshooting (#97305) (#97490) Discovery, like cluster membership, can also be affected by network-like issues (e.g. GC/VM pauses, dropped packets and blocked threads) so this commit duplicates the troubleshooting info across both places. | 10 July 2023, 10:14:55 UTC |
| 4b68b08 | Daniel Mitterdorfer | 10 July 2023, 07:09:40 UTC | [Profiling] Propagate mget failures (#97451) (#97483) The Universal Profiling plugin uses the `mget` API internally to enrich profiling data. If the `mget` API fails to retrieve an item, it will not error but instead provide error details per item. Previously we have skipped such items leading to puzzling results e.g. when a user was lacking proper permissions (user would see an empty page instead of an error). With this commit we abort processing on the first failed item and return an appropriate error response. | 10 July 2023, 07:09:40 UTC |
| f6ebb91 | eyalkoren | 10 July 2023, 06:38:31 UTC | Fix timestamp as object at root level for APM Server (#97401) (#97478) | 10 July 2023, 06:38:31 UTC |
| baf0ff1 | David Turner | 09 July 2023, 20:32:15 UTC | Avoid getStateForMasterService where possible (#97304) (#97479) `Coordinator#getStateForMasterService` will rebuild the state to add a no-master block, which can be rather expensive if there are lots of index blocks to process too. This commit replaces most of its usages with something cheaper. Relates #97295 | 09 July 2023, 20:32:15 UTC |
| 85d356d | Ryan Ernst | 07 July 2023, 23:02:43 UTC | Clarify keystore add-file command behavior (#97474) (#97475) This commit adds a little more explanation to what add-file does on the keystore. relates #97459 | 07 July 2023, 23:02:43 UTC |
| 126ea86 | Mayya Sharipova | 07 July 2023, 20:08:39 UTC | Add vector search file extensions to preload info (#97375) (#97471) Add vector search file extensions to preload info Related to #96955 | 07 July 2023, 20:08:39 UTC |
| 0dd60e5 | James Rodewig | 07 July 2023, 15:28:00 UTC | [8.9] [DOCS] Update Elastic GeoIP service link (#97455) (#97456) | 07 July 2023, 15:28:00 UTC |
| a504a30 | James Rodewig | 07 July 2023, 13:50:58 UTC | [8.9] [DOCS] Add TOC to landing page (#97437) (#97447) Co-authored-by: Abdon Pijpelink <abdon.pijpelink@elastic.co> | 07 July 2023, 13:50:58 UTC |
| fe7cda3 | Pooya Salehi | 07 July 2023, 13:23:35 UTC | Do not use auto_expand_replica in mget yaml rest tests (#97427) (#97445) As described in the issue, the change in #96763 has made the MixedClusterClientYamlTestSuiteIT for mget fail very often. For now, let's take the same approach that we have for get. Closes #97236 | 07 July 2023, 13:23:35 UTC |
| bee3f35 | István Zoltán Szabó | 07 July 2023, 08:03:07 UTC | [8.9][DOCS] Adds release highlight about text embedding for kNN is GA. (#97420) | 07 July 2023, 08:03:07 UTC |
| 76630aa | Mary Gouseti | 07 July 2023, 07:37:33 UTC | Change the way feature flag is used for the data stream lifecycle telemetry. (#97425) | 07 July 2023, 07:37:33 UTC |
| 3c5fd25 | Martijn van Groningen | 06 July 2023, 12:04:36 UTC | [8.9] Fix mapping parsing logic to determine synthetic source is active. (#97355) (#97418) * Fix mapping parsing logic to determine synthetic source is active. (#97355) Take index mode into account during parsing of the mapping when determining whether source is synthetic Fixes #97320 * update skip version | 06 July 2023, 12:04:36 UTC |
| 26ca704 | Kostas Krikellas | 06 July 2023, 09:14:37 UTC | Simplify TDigestTests.testFewValues (#97407) (#97414) TDigests are known to be inaccurate when they only contain a handful of singletons. | 06 July 2023, 09:14:37 UTC |
| bd62e14 | Craig Taverner | 05 July 2023, 16:13:29 UTC | Documentation for time-series geo_line (#97373) (#97404) * Documentation for time-series geo_line * Fix incorrect ids in geoline docs * Some updates from review Added image of kibana map, improved first example, linked to TSDS and added section on line simplification with link to wikipedia. * Diagrams of truncation versus simplification | 05 July 2023, 16:13:29 UTC |
| 4bf5448 | Chris Hegarty | 05 July 2023, 09:28:07 UTC | Update plugin cli/scanner to ASM 9.5 (#97385) This commit updates the plugin cli and scanner components to use ASM 9.5. The update is required to successfully test with JDK 21. Tests in this component programatically run the java source compiler, which generates class files with major version 65, then tries to parse those generated class files. Without this change the tests fail with java.lang.IllegalArgumentException: Unsupported class file major version 65. | 05 July 2023, 09:28:07 UTC |
| 1374fab | Abdon Pijpelink | 05 July 2023, 08:30:34 UTC | [DOCS] Add note about Kibana warning in ILM tutorial (#97346) (#97383) | 05 July 2023, 08:30:34 UTC |
| 740e8cd | David Kyle | 04 July 2023, 16:00:13 UTC | [ML] Fix WordPiece tokenization where stripping accents results in an empty string (#97354) (#97367) Stripping accents sometimes results in an empty string which is then skipped by the WordPiece function causing an error. The basic tokenizer now consumes tokens until a non empty string is found | 04 July 2023, 16:00:13 UTC |
| d8a2999 | Przemyslaw Gomulka | 04 July 2023, 15:32:14 UTC | Initialise ES logging in CLI (#97353) (#97366) ES logging has to be explicitly initialised with a call to LoggerFactory.setInstance This is normally done as LogConfigurator.configure(Environment,boolean) is calling this as part of initPhase1 However CLI tools were not using that method. Cli tools are using LogConfigurator.configureWithoutConfig and that method was not setting up ES logging This commit modifies configureWithoutConfig to also configure esLogging It also adds a LogConfigurator.configureESLogging call explicitly in CliToolLauncher This allows to build a new cli-launcher.jar and replace it in previous 8.7-8.9 clusters closes #97350 | 04 July 2023, 15:32:14 UTC |
| 5663098 | Tanguy Leroux | 04 July 2023, 09:05:38 UTC | Fix reused/recovered bytes for files that are recovered from cache (#97278) (#97347) Fix in #95987 wasn't complete, I did not restore all the previous behavior. Closes #95994 | 04 July 2023, 09:05:38 UTC |
| 0c0571a | Daniel Mitterdorfer | 04 July 2023, 08:10:18 UTC | [Profiling] Wait for ILM policies to be created (#97341) (#97342) In order to create profiling indices we rely on index templates, hence we defer index creation until all of them are present. Some of these index templates might also specify an ILM policy. However, we did not defer index creation until the ILM policy was present which can lead to races between index and ILM policy creation. In this commit we add an additional check to wait until all required ILM policies have been created. Only then we proceed to create indices. | 04 July 2023, 08:10:18 UTC |
| 3febf8c | Nhat Nguyen | 03 July 2023, 21:31:06 UTC | Fix delay timeout in MockTransportService (#97336) (#97337) We can pass an invalid duration to TimeValue, which accepts either -1 or an non-negative duration. | 03 July 2023, 21:31:06 UTC |
| 2500484 | Mayya Sharipova | 03 July 2023, 19:33:54 UTC | Add synonyms feature for test clusters (#97334) For snapshots builds we automatically enable all feature flags, but for release builds they need to be explicitly added to test clusters for tests. This PR does it for synonyms feature. Closes #96641, #97177 | 03 July 2023, 19:37:58 UTC |
| d370ca4 | Pablo Alcantar Morales | 03 July 2023, 15:23:06 UTC | [bugfix] Ensures that we're keeping up with cluster settings in HealthMetadataService (#97135) (#97327) Refactor the code to ensure that the HealthMetadataService keeps up with the settings changes all the time (not only when the cluster is non-blocked) and post the changes to the ClusterState once the cluster is stable. Before we were discarding changes to some settings in case the cluster was not stable (e.g..: not all nodes in the same version), this provoked that if a user changed a setting, once a new master was elected, the internal HealthMetadata object was reset to the default values, do not respecting new settings values. With this change, we hold an in-memory representation of the HealthMetadata within HealthMetadataService which is always up-to-date with the latest user settings and we publish to the ClusterState only when is viable to publish. | 03 July 2023, 15:23:06 UTC |
| f3f8fae | David Turner | 03 July 2023, 14:58:01 UTC | Handle failure in TransportUpdateAction#handleUpdateFailureWithRetry (#97290) (#97325) Here executor(request.getShardId()) may throw, but we're already handling a failure so we cannot simply let this exception bubble up. This commit adjusts things to catch the exception, using it to fail the listener. Closes #97286 | 03 July 2023, 14:58:01 UTC |
| 101aafe | Andrei Dan | 03 July 2023, 09:43:14 UTC | Chunk the GET _ilm/policy response (#97251) (#97316) Chunk the GET _ilm/policy response | 03 July 2023, 09:43:14 UTC |
| 998c77b | Lee Hinman | 30 June 2023, 19:39:07 UTC | [8.9] Bump Guava from 30.1-jre to 32.0.1-jre (#97211) (#97254) * Bump Guava from 30.1-jre to 32.0.1-jre (#97211) This bumps Watcher's Guava dependency to 32.0.1. * Regenerate guava checksums for Gradle --------- Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> | 30 June 2023, 19:39:07 UTC |
| cacc0cc | Lee Hinman | 30 June 2023, 15:20:13 UTC | Move get lifecycle API to Management thread pool and make cancellable (#97248) (#97284) * Move get lifecycle API to Management thread pool and make cancellable This commit moves the `TransportGetLifecycleAction` to execute on the `Management` thread pool, and to be cancellable so that wildcard requests looping through all or a subset of policies checks for cancellation while running. Resolves #96568 | 30 June 2023, 15:20:13 UTC |
| 75cb6fe | Hendrik Muhs | 30 June 2023, 11:00:04 UTC | [ML] improve model downloader robustness (#97274) (#97276) use the resolver from ModelLoaderUtils, fix problems finding the vocabulary file | 30 June 2023, 11:00:04 UTC |
| 249c033 | James Rodewig | 29 June 2023, 21:12:13 UTC | [DOCS] Note license requirements for CCS (#97252) (#97256) Notes that CCS requires both clusters to use the same license level for full capabilities. (cherry picked from commit ff84ad146903d4bf69b3bae257e4a602300c7dcd) | 29 June 2023, 21:12:13 UTC |
| c0865f0 | Mark Vieira | 29 June 2023, 20:27:39 UTC | Bump versions after 7.17.11 release | 29 June 2023, 20:30:37 UTC |
| 40041ae | Mark Vieira | 29 June 2023, 19:28:34 UTC | Prune changelogs after 8.8.2 release | 29 June 2023, 19:28:34 UTC |
| 999946c | Mark Vieira | 29 June 2023, 19:24:34 UTC | Bump versions after 8.8.2 release | 29 June 2023, 19:27:52 UTC |
| 565214f | Mary Gouseti | 29 June 2023, 15:58:58 UTC | Rename origin text from lifecycle to data_stream_lifecycle (#97230) (#97241) | 29 June 2023, 15:58:58 UTC |
| 9ff6da9 | Abdon Pijpelink | 29 June 2023, 13:39:05 UTC | [DOCS] Fix typo (#97233) (#97239) | 29 June 2023, 13:39:05 UTC |
| 1efb8a4 | Daniel Mitterdorfer | 29 June 2023, 12:14:52 UTC | [Profiling] Don't force-merge indices (#97229) (#97232) With this commit we remove the `force-merge` action from the default ILM policy for Universal Profiling. We do this because our experiments have shown that `force-merge` only leads to a marginal reduction in storage space (around 2% in total) while having potentially negative side-effects: * It is I/O and CPU intensive leading to potential stress on the warm tier. * The required disk space temporarily increases up to double size of the index being force-merged. Therefore we stop doing force merges by default. | 29 June 2023, 12:14:52 UTC |
| 2db2f58 | Slobodan Adamović | 29 June 2023, 08:08:37 UTC | [DOC] Consistently reference indices parameter in API key examples (#97174) (#97221) This PR adjusts API key examples to consistently use indices instead of index parameter. Technically, both are supported as request parameters, but indices is more aligned with get API key response. Relates to #97027, where the change was already made for create API key examples. | 29 June 2023, 08:08:37 UTC |
| cd14c21 | Mary Gouseti | 29 June 2023, 06:39:03 UTC | Start with data stream lifecycle documentation (#95326) (#97189) | 29 June 2023, 06:39:03 UTC |
| 9a17eb3 | Salvatore Campagna | 28 June 2023, 16:03:33 UTC | Include release highlight for query rewrite (#97178) (#97202) Update the changelog after making the original PR a `release highlight`. See #96161 | 28 June 2023, 16:03:33 UTC |
| 2f12a52 | Slobodan Adamović | 28 June 2023, 16:00:54 UTC | [Test] Use real instance of ReservedRolesStore to avoid NPE (#97185) (#97201) Since ReservedRolesStore must be properly initialized before calling real methods, it is not possible to mock it and then still call real methods. Given that there is no real need for CompositeRolesStoreTests to use mocked ReservedRolesStore, this PR changes the test setup to use real instance and thus avoid NPE when accessing RESERVED_ROLES property. Fixes #97172 | 28 June 2023, 16:00:54 UTC |
| 4570e83 | Christoph Büscher | 28 June 2023, 15:11:40 UTC | Mute some knn yml tests (#97191) (#97197) | 28 June 2023, 15:11:40 UTC |
| f363a0d | Jack Conradson | 28 June 2023, 14:19:31 UTC | Add basic documentation for sub searches (#97025) (#97195) This adds basic documentation for the sub_searches top-level element in the search API. (#96224). | 28 June 2023, 14:19:31 UTC |
| 3567ac8 | Abdon Pijpelink | 28 June 2023, 14:14:40 UTC | [DOC] Fixes a link in nori_part_of_speech token filter (#97158) (#97192) * [DOCS] Fixes a link in nori_part_of_speech token filter * [DOCS] Fixes a typo in the link --------- Co-authored-by: Wonseop Kim <refarde@kakao.com> Co-authored-by: Wonseop Kim <refarde@gmail.com> | 28 June 2023, 14:14:40 UTC |
| adc754b | Kostas Krikellas | 28 June 2023, 13:32:59 UTC | [TEST] Increase error bound in flaky test (#97176) (#97186) * Use a highly accurate TDigest to compute MAD. MedianAbsoluteDeviation builds a second TDigest over the centroids of the TDigest that's computed over the input. The second TDigest should have only a handful of values, so using an AVLTreeDigest produces more accurate results with no significant performance penalty. Note that the HybridDigest can't use SortingDigest internally for large inputs, even if the number of centroids is fairly small (likely less than 1000). * Update docs/changelog/97176.yaml * Increase error in test. * Delete docs/changelog/97176.yaml | 28 June 2023, 13:32:59 UTC |
| 53c6a6f | Iraklis Psaroudakis | 28 June 2023, 08:19:06 UTC | Fix testScheduledRefresh (#97147) (#97160) Fixes #96920 | 28 June 2023, 08:19:06 UTC |
| 04c4738 | Adrien Grand | 27 June 2023, 17:22:57 UTC | Upgrade to Lucene 9.7.0. (#97073) (#97156) This is the official release, and is very close to the snapshot we're currently on. | 27 June 2023, 17:22:57 UTC |
| 69a089e | Tanguy Leroux | 27 June 2023, 15:24:15 UTC | Fix cluster settings update task acknowledgment (#97111) (#97148) ClusterUpdateSettingsTask uses a "changed" variable to set the value of the acknowledged flag of the ClusterUpdateSettingsResponse that is returned to the client. The "changed" variable is computed from the cluster state update but there is no guarantee that the update will be acknowledged. In case the update failed to be acked (onAckFailure) it should return the flag appropriately. | 27 June 2023, 15:24:15 UTC |
| b811d11 | Slobodan Adamović | 27 June 2023, 14:30:27 UTC | [DOC] Document role restriction for API keys (#97027) (#97145) This PR adds API documentation for role restriction and workflows which are introduced in #96744 and #96215. --------- Co-authored-by: Abdon Pijpelink <abdon.pijpelink@elastic.co> Co-authored-by: Yang Wang <ywangd@gmail.com> | 27 June 2023, 14:30:27 UTC |
| 90a3c5c | Pablo Alcantar Morales | 27 June 2023, 09:11:05 UTC | Document the enhancements to ILM Health Indicator (#96980) (#97131) | 27 June 2023, 09:11:05 UTC |
| 4d3b886 | Ryan Ernst | 26 June 2023, 21:49:39 UTC | Capture max processors in static init (#97119) (#97122) The number of processors available to the jvm can change over time. However, most of Elasticsearch assumes this value is constant. Although we could rework all code relying on the number of processors to dynamically support updates and poll the jvm, doing so has little value since the processors changing is an edge case. Instead, this commit fixes validation of the node.processors setting (our internal number of processors) to validate based on the max processors available at launch. closes #97088 | 26 June 2023, 21:49:39 UTC |
| bdd177c | Iraklis Psaroudakis | 26 June 2023, 14:25:19 UTC | Upgrade Netty to 4.1.94.Final (#97040) (#97112) Just staying up to date. As requested, backported from https://github.com/elastic/elasticsearch/pull/97040 | 26 June 2023, 14:25:19 UTC |
| e2961c1 | Kostas Krikellas | 26 June 2023, 13:11:10 UTC | Skip `SortingDigest` when merging a large digest in `HybridDigest` (#97099) (#97106) * Skip SortingDigest when merging a large digest in HybridDigest. This is a small performance optimization that avoids creating an intermediate SortingDigest when merging a digest tracking many samples. The current behavior is to keep adding values to SortingDigest until we cross the threshold for switching to MergingDigest, at which point we copy all values from SortingDigest to MergingDigest and release the former. As a side cleanup, remove the methods for adding a list of digests. It's not used anywhere and it can be tricky to get right - the current implementation for HybridDigest is buggy. * Update docs/changelog/97099.yaml | 26 June 2023, 13:11:10 UTC |
| 4cf476e | Tanguy Leroux | 26 June 2023, 12:35:49 UTC | Fix ESCCRRestTestCase.verifyAutoFollowMonitoring (#97054) (#97104) The method `verifyAutoFollowMonitoring` searches for `ccr_auto_follow_stats` documents in `monitoring-es-*` indices to see if one document has the field `number_of_successful_follow_indices` greater than 0. If such document is found, it means that x-pack monitoring successfully indexed documents about CCR stats and that CCR at that time had at least 1 following index successfully running. But the current `verifyAutoFollowMonitoring` method only check the first 10 docs returned by the search requests. Recent failures of `AutoFollowIT.testAutoFollowPatterns` on 7.17 branch indicates that more than 10 docs were found, so I suspect one of them has `number_of_successful_follow_indices` > 0 but is not returned in the first 10 top docs. This pull request changes the method to only count documents with `number_of_successful_follow_indices` > 0. Closes #91984 | 26 June 2023, 12:35:49 UTC |
| 4b72136 | Tanguy Leroux | 26 June 2023, 12:15:30 UTC | Fix reused/recovered bytes for files that are only partially recovered from cache (#95987) (#97102) Before #95891 a file was considered as reused in recovery if it was fully cached in the cold persistent cache. Otherwise the full file length was reported as recovered from the blob store during prewarming. While working on #95891 I changed the CachedBlobContainerIndexInput.prefetchPart() method to be more precise on the number of bytes that were effectively read from the blob store during prewarming. But this obviously broke some tests (#95970) because a file cannot be partially recovered from disk and from remote. This change restores the previous behavior with one adjustment: the file is considered as reused if the prewarm method effectively prefetched 0 bytes. Closes #95970 Closes #95994 | 26 June 2023, 12:15:30 UTC |
| 018e362 | Valeriy Khakhutskyy | 26 June 2023, 11:30:36 UTC | [TEST] Mute RemoteClusterClientTests.testConnectAndExecuteRequest #97084 (#97100) Relates #97080 | 26 June 2023, 11:30:36 UTC |
| fa4e495 | Andrei Dan | 26 June 2023, 10:24:39 UTC | Limit the details field length we store for each SLM invocation (#97038) (#97094) | 26 June 2023, 10:24:39 UTC |
| 1e9b4e7 | István Zoltán Szabó | 26 June 2023, 09:54:44 UTC | [DOCS] Adds API docs for bert_ja text embedding tokenizer option (#96873) (#97095) | 26 June 2023, 09:54:44 UTC |
| 07f15a3 | István Zoltán Szabó | 26 June 2023, 09:33:53 UTC | [DOCS] Adjusts the note about minimum recommended node size on the ELSER tutorial page (#97083) (#97089) Co-authored-by: David Roberts <dave.roberts@elastic.co> | 26 June 2023, 09:33:53 UTC |
| 88af9cc | Liam Thompson | 26 June 2023, 08:18:41 UTC | [DOCS] Delete duplicated link (#97076) (#97078) | 26 June 2023, 08:18:41 UTC |
| 2f8efae | Abdon Pijpelink | 24 June 2023, 10:20:07 UTC | [DOCS] Add link to Render Search Application Query docs (#97067) (#97068) (cherry picked from commit 012dea54dc8707c17abcc371a992cca9fec28c8f) | 24 June 2023, 10:20:07 UTC |
| f0f6bb0 | Lisa Cawley | 24 June 2023, 08:10:33 UTC | [DOCS] Placeholder for synonym APIs (#97063) (#97065) (cherry picked from commit af6cd5ff84cc171a552901cb2f9e810e759f0838) | 24 June 2023, 08:10:33 UTC |
| ee521ec | Ryan Ernst | 24 June 2023, 02:15:56 UTC | Fix Painless method lookup over unknown super interfaces (#97062) (#97064) In Java 21 List now extends SequencedCollection, instead of Collection directly. When resolving methods Painless starts at the defined type, and iterates up through super classes and interfaces. Unfortunately if a superinterface was not known, as it is for SequencedCollection since it is not in the allowed list of classes, method resolution would give up. This commit adjusts the superinterface interation to continue traversing until the method is found or no more superinterfaces are found. fixes #97022 | 24 June 2023, 02:15:56 UTC |
| 652a2e6 | Yang Wang | 23 June 2023, 14:00:37 UTC | [Test] Initialize ReservedRolesStore for each test method (#97031) (#97048) This PR changes the initialization of ReservedRolesStore from once per test class to once per test method so that it is in the expected state for each test. The failure was because individual test can instantiate ReservedRolesStore differently which in turn mutates the class wide shared static field of reserved-roles. Resolves: #97017 | 23 June 2023, 14:00:37 UTC |
