5b38441 | Mark Vieira | 10 December 2021, 23:53:32 UTC | Patch log4j JAR to remove JndiLookup class (#81631) | 10 December 2021, 23:53:32 UTC |
cdc5ae4 | Chris Hegarty | 10 December 2021, 17:47:49 UTC | Add logging property (#81624) | 10 December 2021, 17:47:49 UTC |
c0c5720 | James Rodewig | 09 December 2021, 23:20:20 UTC | [DOCS] Clarify when `index` parameter is required for Watcher index action (#81566) (#81610) If you dynamically add an `_index` value, the `index` parameter is not required for the Watcher index action. Co-authored-by: James Rodewig <40268737+jrodewig@users.noreply.github.com> # Conflicts: # x-pack/docs/en/watcher/actions/index.asciidoc Co-authored-by: Daisuke Harada <1519063+dharada@users.noreply.github.com> | 09 December 2021, 23:20:20 UTC |
1674ce8 | William Brafford | 09 December 2021, 22:41:29 UTC | Set correct version bounds for system features rest API tests (#81465) (#81607) * Use correct version bounds for API test * Use correct version bounds for migration API test Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> | 09 December 2021, 22:41:29 UTC |
d8df234 | Rory Hunter | 09 December 2021, 21:20:31 UTC | Remove Cloud images from 7.16 (#80742) We don't actually need Cloud images before v8.0, so remove all references and supporting code. | 09 December 2021, 21:20:31 UTC |
08bee54 | Rene Groeschke | 09 December 2021, 18:53:50 UTC | Ignore InternalTestRerunPluginFuncTest (#81591) Ignore these tests as we currently do not use the plugin due to it's flakiness based on how things are implemented in gradle | 09 December 2021, 18:53:50 UTC |
b57ffab | Seth Michael Larson | 09 December 2021, 17:36:09 UTC | Add optional body to ml.open_job and ml.forecast APIs (#81594) | 09 December 2021, 17:36:09 UTC |
17c211b | James Rodewig | 09 December 2021, 14:20:00 UTC | [DOCS] Clarify rolling upgrade support for minor versions (#81444) (#81588) We only support rolling upgrades between minor versions of the same major. It also adds some tags to reuse duplicated content. # Conflicts: # docs/reference/upgrade/rolling_upgrade.asciidoc | 09 December 2021, 14:20:00 UTC |
850aa05 | Przemyslaw Gomulka | 09 December 2021, 10:16:08 UTC | Emit x-elastic-product-origin into logs backport(#81381) (#81518) This commit emits the value of x-elastic-product-origin header into elasticsearch.elastic_product_origin field in ES deprecation logs and indexed deprecation logs. x-elastic-product-origin header is intended to identify the elastic stack origin and allow to ignore deprecations emitted by the stack. ES v7 logging differs from v8 as it creates the JSON layout on its own. The additional custom field have to be declared in log4j.properties and ECSLayout used for indexing deprecated logs also has to be updated backports #81381 | 09 December 2021, 10:16:08 UTC |
f4b7127 | Hendrik Muhs | 09 December 2021, 09:06:26 UTC | [Transform] Repair TransformSurvivesUpgradeIT for 7.16->7.16/17 upgrade (#81571) don't check for audit template after 7.16 fixes #79230 | 09 December 2021, 09:06:26 UTC |
d85dd2f | David Turner | 09 December 2021, 09:04:00 UTC | Oversharding is also indices and fields (#81511) (#81574) Today the _Size your shards_ docs focus on shard size and count, but in fact index count and field count are also important. This commit expands these docs a bit to cover this observation too. | 09 December 2021, 09:04:00 UTC |
6bef861 | Przemyslaw Gomulka | 09 December 2021, 07:54:36 UTC | Add trace.id to indexed deprecation logs (#81524) #75331 missed adding trace.id to indexed deprecation logs (these use ECSJsonLayout instead of ESJsonLayout). | 09 December 2021, 07:54:36 UTC |
48cbeda | James Baiera | 08 December 2021, 23:57:59 UTC | Deprecate multifield mappings templates (#81329) Using fields within fields in mappings is deprecated, but previously would not be listed in the deprecation API if the mapping was part of an index template, a component template, or within mapping dynamic_template sections which can be present in either of the locations. This change checks all six combinations of locations for the deprecated mapping structure. | 08 December 2021, 23:57:59 UTC |
7d8f0b3 | Stuart Tettemer | 08 December 2021, 23:31:43 UTC | [7.16] Script: Implicit context cache (#81552) (#81560) * [7.16] Script: Implicit context cache (#81552) The script context cache is deprecated. In 7.16 the default value of `script.max_compilation_rate` was switched from `"use-context"`, to `"150/5m". That means uses would have to explicitly set "use-context" to use any of the context cache families of settings: `script.context.*.max_compilations_rate` `script.context.*.cache_max_size` `script.context.*.cache_expire` On upgrades to 7.16, if the customer was using the default and had set any of those settings without setting `script.max_compilation_rate: "use-context"`, the upgrade would be rejected. To avoid an unintentional breaking change, the script service will now **implicitly** use the script context cache if `script.max_compilation_rate` is **unset** and any of the context cache family of settings is set. The context cache will also be used if `script.max_compilation_rate: "use-context"`, as before. Fixes: #81486 Backport: ba6c17d | 08 December 2021, 23:31:43 UTC |
1cd2640 | Mark Vieira | 08 December 2021, 20:30:09 UTC | Ignore temporary files when syncing test cluster distribution (#81482) (#81555) | 08 December 2021, 20:30:09 UTC |
8ee0e59 | Hendrik Muhs | 08 December 2021, 18:02:07 UTC | fix test randomization issue when randomization turned of pit (#81540) (#81544) fix test randomization issue when randomization turned off pit fixes #81537 | 08 December 2021, 18:02:07 UTC |
b74888c | James Rodewig | 08 December 2021, 17:53:41 UTC | [DOCS] Update field capabilities API mentions in RNs (#81541) Edits the 7.16 and 7.14 release notes to include "field capabilities API" in items related to the API. This makes it easy for users to find the items using an on-page search. | 08 December 2021, 17:53:41 UTC |
847309f | Gilad Gal | 08 December 2021, 14:41:31 UTC | Update search-across-clusters.asciidoc Vector tiles search moved to GA in 7.16 | 08 December 2021, 14:41:31 UTC |
0916612 | Chris Hegarty | 08 December 2021, 12:55:11 UTC | Allow plugins MBeanTrustPermission (#81508) (#81516) MBeans are sometimes used by third party libraries, e.g. to report metrics through JMX. This commit builds upon the initial set of MBean permissions added in #76329, to further alllow a plugin register an MBean, i.e to allow plugins to grant MBeanTrustPermission("register"). | 08 December 2021, 12:55:11 UTC |
cc91624 | Bogdan Pintea | 08 December 2021, 12:51:11 UTC | SQL: Only resolve aliases if asked for local cluster (#81158) (#81517) This fixes an issue where the local aliases are returned also in the case the cluster patern indicates a remote cluster. The alias resolution is now guarded by a cluster pattern match. | 08 December 2021, 12:51:11 UTC |
7250941 | István Zoltán Szabó | 08 December 2021, 11:56:47 UTC | [7.16][DOCS] Changes HLRC Client links. (#81513) | 08 December 2021, 11:56:47 UTC |
eeb39f8 | Andrei Stefan | 08 December 2021, 11:56:25 UTC | EQL: HLRC documentation (#80979) | 08 December 2021, 11:56:25 UTC |
3798ff6 | Andrei Dan | 08 December 2021, 09:12:07 UTC | Allow the `unfollow` action in the frozen phase (#81434) (#81495) * Allow the `unfollow` action in the frozen phase The `unfollow` actions is injected before some of the actions that yield the managed index unsafe to use in a CCR environment (eg. rollover, searchable_snapshot, shrink). This was not allowed in the `frozen` phase, however we do have the `searchable_snapshot` action available. This commit makes the `unfollow` action available in the frozen phase. (cherry picked from commit 8ba6f55e0a7b3751355eb7e52f17a9cb80f2b7f7) Signed-off-by: Andrei Dan <andrei.dan@elastic.co> # Conflicts: # x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ilm/TimeseriesLifecycleTypeTests.java | 08 December 2021, 09:12:07 UTC |
969e565 | Hendrik Muhs | 08 December 2021, 08:44:56 UTC | [7.16] [Transform] handle pit index not found error (#81368) (#81491) Do not fail the transform if pit search fails with index not found as a result of an index that got deleted via ILM, if that index is part of a search that selects indices using a wildcard, e.g. logs-*. If pit search fails, the search is retried using search without a pit context. The 2nd search might fail if the source targets an explicit index. In addition the usage of the pit API can not be disabled by transform. fixes elastic#81252 relates elastic#81256 | 08 December 2021, 08:44:56 UTC |
5e0003f | Mat Schaffer | 08 December 2021, 07:06:20 UTC | Allow configuration of ThreadPool warning times (#81487) We added some warn logging in #72465 but did not register the settings that control the thresholds. This commit addresses that oversight. | 08 December 2021, 07:42:25 UTC |
7d493f6 | David Roberts | 08 December 2021, 07:22:11 UTC | [ML] Make model snapshot upgrade autoscaling friendly (#81303) Model snapshot upgrade was not taking autoscaling into account when doing node assignment. Backport of #81123 | 08 December 2021, 07:22:11 UTC |
8d42b0d | James Rodewig | 07 December 2021, 22:09:35 UTC | [DOCS] Fix `combined_fields` query ref in `multi_match` query docs (#81456) (#81472) The current `multi_match` docs contain an erroneous reference to the `combined_fields` query. This updates the reference to reference the correct query. Relates to https://github.com/elastic/elasticsearch/pull/76893 # Conflicts: # docs/reference/query-dsl/combined-fields-query.asciidoc | 07 December 2021, 22:09:35 UTC |
9d8029e | Joe Gallo | 07 December 2021, 21:39:49 UTC | Add an assert to detect double decrementing of ingest metrics (#81450) | 07 December 2021, 21:49:55 UTC |
6626cc9 | Dan Hermann | 07 December 2021, 18:07:03 UTC | Mute failing XPackRestIT test {p0=data_stream/10_basic/...} test (#81463) | 07 December 2021, 18:07:03 UTC |
7d1fc65 | Ievgen Degtiarenko | 07 December 2021, 16:54:43 UTC | Stop auto-followers when metadata is resetted (#81290) (#81452) (cherry picked from commit bdc1b73006434f50aae57ba499ef424f7ffe48aa) | 07 December 2021, 16:54:43 UTC |
abc89b1 | Dan Hermann | 07 December 2021, 15:25:27 UTC | Bump versions for 7.16.0 release | 07 December 2021, 16:17:36 UTC |
ced9d24 | debadair | 07 December 2021, 16:15:56 UTC | Remove coming tag for 7.16. (#81455) | 07 December 2021, 16:15:56 UTC |
1b8c867 | Joe Gallo | 07 December 2021, 16:03:41 UTC | _tier_preference docs changes for 7.16 (#81401) | 07 December 2021, 16:03:41 UTC |
a545e86 | Rene Groeschke | 07 December 2021, 16:02:46 UTC | Update gradle wrapper for plugin examples (#81435) (#81446) | 07 December 2021, 16:02:46 UTC |
6a25058 | Rory Hunter | 07 December 2021, 15:14:33 UTC | Checkstyle shadows vars pt10 (#81361) (#81437) Backport of #81361. Part of #19752. Fix more instances where local variable names were shadowing field names. | 07 December 2021, 15:14:33 UTC |
d0d0e95 | Tanguy Leroux | 07 December 2021, 14:48:26 UTC | Mention prefixes of mounted index in ILM Searchable Snapshot action (#81421) (#81442) The searchable snapshot action mounts snapshots as indices with a different prefix depending of the phase. This commit tries to mention them in the docs. | 07 December 2021, 14:48:26 UTC |
fd8702c | James Rodewig | 07 December 2021, 14:32:56 UTC | [DOCS] Remove outdated APM ifeval (#81402) (#81439) Removes the old `ifeval` for links to 7.x APM documentation. 7.x no longer exists, so this code isn't needed. Originally added in https://github.com/elastic/elasticsearch/pull/56538. Co-authored-by: Brandon Morelli <brandon.morelli@elastic.co> | 07 December 2021, 14:32:56 UTC |
8305dbd | Dan Hermann | 07 December 2021, 13:59:26 UTC | [DOCS] Release notes for v7.16.0 (#81369) | 07 December 2021, 13:59:26 UTC |
c874e70 | Dan Hermann | 07 December 2021, 13:59:07 UTC | [DOCS] Release highlights for 7.16.0 (#81370) | 07 December 2021, 13:59:07 UTC |
d26b202 | Andrei Dan | 07 December 2021, 12:41:00 UTC | [7.16] Implement getMinimalSupportedVersion for all NameDiffs (#81374) (#81424) * Implement getMinimalSupportedVersion for all NameDiffs (#81374) This removes the default implementation of `getMinimalSupportedVersion` and implements it in all `NameDiffs` implementations. The default implementation was rather broad, `minimumIndexCompatibilityVersion` as the index compatibility are wider than our wire compatibility. Having a default implementation also had the implementors forget about adjusting the wire compatibility to the correct version. (cherry picked from commit 5546d5a267d9464d0703b075fbddfd963ee9db0d) Signed-off-by: Andrei Dan <andrei.dan@elastic.co> # Conflicts: # x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/MlMetadata.java # x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/transform/TransformMetadata.java # x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/allocation/TrainedModelAllocationMetadata.java | 07 December 2021, 12:41:00 UTC |
d64e867 | Salvatore Campagna | 07 December 2021, 10:21:16 UTC | [DOCS] Fix the weighed average documentation (#81307) (#81417) The documentations states that if the `weight` field is missing, and no explicit missing configuration is provided, a default value of 1 is used. This is incorrect and does not match the implementation of the weighted average aggregator. In this specific case the document is skipped, instead. | 07 December 2021, 10:21:16 UTC |
5d8ac4b | Gilad Gal | 07 December 2021, 08:24:40 UTC | Update search-across-clusters.asciidoc The QL team added support for SQL via CCS to both 7.16 and 8.0 - https://github.com/elastic/elasticsearch/pull/78903 | 07 December 2021, 08:24:40 UTC |
4a1e55d | James Rodewig | 06 December 2021, 22:46:13 UTC | [DOCS] Fix 7.16 breaking change formatting (#81390) | 06 December 2021, 22:46:13 UTC |
a8dbf17 | David Turner | 06 December 2021, 13:34:12 UTC | Add explicit warning not to touch repository contents (#81295) (#81366) Manipulating the contents of a snapshot repository is a very bad idea, but it turns out we don't call this out in the docs anywhere. This commit adds a warning about this. | 06 December 2021, 13:34:12 UTC |
81ebcbc | Rory Hunter | 06 December 2021, 12:29:39 UTC | Checkstyle shadows vars pt9 (#81296) (#81359) Part of #19752. Fix more instances where local variable names were shadowing field names. | 06 December 2021, 12:29:39 UTC |
77a65eb | Keith Massey | 03 December 2021, 22:18:49 UTC | Adding a deprecation info API check for discovery.zen.minimum_master_nodes (#81282) This commit adds deprecation info API checks for the 15 or so discovery.zen settings that did not already have deprecation info API checks. Closes #81260 | 03 December 2021, 22:18:49 UTC |
512b60c | James Baiera | 03 December 2021, 20:35:35 UTC | Monitoring migration guide updates (#79756) Adds additional documentation for the deprecation of the Monitoring plugin collection and shipping settings. | 03 December 2021, 20:35:35 UTC |
d9f8fca | Rory Hunter | 03 December 2021, 19:30:53 UTC | Revert some EQL shadowed var changes and improve HiddenField rule (#81317) Change our fork of `HiddenFieldCheck` rule to allow it to ignore shadowed variables for methods shorter than a configured minimim line count. While this technically defeats the point of the rule, in practice short methods are easier to verify by eye, compared to longer methods where accidentally referencing the wrong variable is harder to spot. Following this, revert some changes in the EQL code which were required to pass the `HiddenField` check before the `minLineCount` change. Additionally, introduce a mechanism for ignoring shadowed variables when used created a new object. This allowed us to ignore variables in usages of e.g. `ConstructingObjectParser` whose purpose, as the name suggests, is to build objects, and object shadows variables in the course of building objects. | 03 December 2021, 19:36:15 UTC |
4eb9667 | Rory Hunter | 03 December 2021, 19:24:56 UTC | Revert "Use Cloudflare's zlib in Docker images (#81245)" This reverts commit 6582acfc00029213da35a3bc4cee957b015f4210. | 03 December 2021, 19:24:56 UTC |
d68c756 | Lisa Cawley | 03 December 2021, 18:42:35 UTC | [7.16][DOCS] Move trained model APIs out of dataframe analytics (#81315) (#81324) | 03 December 2021, 18:42:35 UTC |
14e1f00 | David Turner | 03 December 2021, 16:42:57 UTC | Fix channel open trace logging (#81308) (#81314) Today we log `Tcp transport channel opened` when the channel is created, before it is connected, which means its addresses are not configured so it appears as `Netty4TcpChannel{localAddress=null, remoteAddress=null}` which is pretty useless. This commit defers the logging until the channel is actually open at which point its addresses are available. | 03 December 2021, 16:42:57 UTC |
34fb605 | Dan Hermann | 03 December 2021, 14:53:11 UTC | Increase test timeout for CoordinatorTests testAllSearchesExecuted (#81281) | 03 December 2021, 14:53:11 UTC |
6582acf | Rory Hunter | 03 December 2021, 09:48:45 UTC | Use Cloudflare's zlib in Docker images (#81245) Closes #81208. Elasticsearch uses zlib for two purposes: * Compression of stored fields with `index.codec: best_compression`, which we use for observability and security data. * Request / response compression. Historically, zlib was packaged within the JDK, so that users wouldn't have to have zlib installed for basic usage of Java. However, the original zlib optimizes for portability and misses a number of important optimizations such as leveraging vectorization support for x86 and ARM architectures. Several forks have been created in order to address this. Since version 9, the JDK uses the system's zlib when available and falls back to the zlib that is packaged within the JDK if a system zlib cannot be found. This commit changes the Docker image to install the Cloudflare fork of zlib, and run Java using the fork instead of the original zlib, so that users of the Docker image can get better performance. Other ES distribution types are out-of-scope, since configuring the JVM to use an alternative zlib requires an environment config as well as installed another zlib, and Docker is the only distribution type where we can control both. | 03 December 2021, 11:34:08 UTC |
130f508 | David Roberts | 03 December 2021, 11:18:58 UTC | [Transform] Fix wire compatibility version for transform cluster state (#81300) Since 7.13 transforms has had a very small custom cluster state that records whether a feature reset is in progress. This should not be sent to older nodes when the cluster state is serialized. Fixes #81292 | 03 December 2021, 11:18:58 UTC |
3b56aee | Yannick Welsch | 02 December 2021, 15:00:50 UTC | Fail fast remote cluster requests (#80589) When using the skip_unavailable option on remote clusters, searches and similar actions should fail quickly instead of waiting on connections to be reestablished. This commit makes it so that search, field caps, resolve index, and term enum cross-cluster requests fail fast when skip_unavailable is set to true on the given cluster. Internal cross-cluster requests (e.g. license checks, CCR, etc.) still continue waiting on connection attempts. Closes #74773 | 03 December 2021, 09:56:37 UTC |
ebd6454 | Rene Groeschke | 03 December 2021, 08:44:05 UTC | Update gradle version to bugfix version 7.3.1 (#81243) (#81288) | 03 December 2021, 08:44:05 UTC |
2ba3632 | Lukas Wegmann | 03 December 2021, 08:23:25 UTC | SQL: Fix rearranging columns in PIVOT queries (#81032) (#81287) Resolves #80952 and also fixes arbitrary rearrangements of pivoted columns (dropping, duplicating and reordering). The bug was caused because fields reordering in `ProjectExec` wrapping `PivotExec` have been ignored. As a result the field mask (from `QueryContainer.columnMask()`) is created wrongly because the attributes in `EsQueryExec.output()` have a different order in `EsQueryExec.output()` as in `QueryContainer.fields()`. | 03 December 2021, 08:23:25 UTC |
66daa10 | Lisa Cawley | 03 December 2021, 02:35:44 UTC | [DOCS] Adds missing timeout parameter to transform APIs (#81129) (#81278) | 03 December 2021, 02:35:44 UTC |
964ece6 | James Rodewig | 02 December 2021, 19:17:44 UTC | [DOCS] Fix syntax error in bulk `dynamic_templates` docs (#81264) (#81267) | 02 December 2021, 19:17:44 UTC |
b1130cf | Yannick Welsch | 02 December 2021, 16:25:23 UTC | Revert "Fail fast remote cluster requests (#80589)" This reverts commit 8ce950b96cc13e0391a092a8e42a9bfd3ab1fd27. | 02 December 2021, 16:25:23 UTC |
8ce950b | Yannick Welsch | 02 December 2021, 15:00:50 UTC | Fail fast remote cluster requests (#80589) When using the skip_unavailable option on remote clusters, searches and similar actions should fail quickly instead of waiting on connections to be reestablished. This commit makes it so that search, field caps, resolve index, and term enum cross-cluster requests fail fast when skip_unavailable is set to true on the given cluster. Internal cross-cluster requests (e.g. license checks, CCR, etc.) still continue waiting on connection attempts. Closes #74773 | 02 December 2021, 15:59:05 UTC |
6fc8166 | Rory Hunter | 02 December 2021, 13:04:26 UTC | Checkstyle shadows vars pt8 (#81147) (#81250) Part of #19752. Fix more instances where local variable names were shadowing field names. Also expand the possible method names that are skipped when checking for shadowed vars, and allow shadowed vars in builder classes. | 02 December 2021, 13:04:26 UTC |
c23f8f6 | Rory Hunter | 02 December 2021, 12:53:12 UTC | Fix shadowed vars pt7 (#80996) (#81144) Part of #19752. Fix more instances where local variable names were shadowing field names. Also modify our fork of HiddenFieldCheck to add the ignoreConstructorBody and ignoredMethodNames parameters, so that the check can ignore more matches. Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> | 02 December 2021, 12:53:12 UTC |
45241aa | Ignacio Vera | 02 December 2021, 08:55:28 UTC | Fix bug filtering collinear points in polygon decomposition (#81155) (#81241) | 02 December 2021, 08:55:28 UTC |
226091f | Benjamin Trent | 01 December 2021, 16:53:21 UTC | [7.16] [ML] fix autoscaling capacity consistency (#81181) (#81221) * [ML] fix autoscaling capacity consistency (#81181) In certain scenarios, it is possible to request a scale up but, a subsequent call, with no memory changes, a scale down would be requested. What this would look like in practice is many subsequent autoscaling actions taking place. It stems from the following situation: Scale up request is asked, a waiting job just barely doesn't fit at the current scale and tips it over into the "next scaling tier" When calculating if scale down is possible, the calculation of the JVM size when compared to the required native size is too small Scale down assumes a smaller node could work, but actually the true JVM percentage wouldn't allow it. This commit also adds two larger situational tests: One specific edge case that failed before this change A more random test that verifies that when a scale_up is asked and all jobs assigned, we do NOT ask for a scale down to a lower tier accidentally * Fixing backport * fixing backport | 01 December 2021, 16:53:21 UTC |
aae6b16 | Martijn van Groningen | 01 December 2021, 15:47:13 UTC | Unmute 'Create data stream' yaml rest test and (#81201) (#81219) move the assertions of replicated field to a new test with different skip versions. (The replicated was recently added and when the 'Create data stream' yaml rest test is ran against older clusters this caused failured) Closes #81164 | 01 December 2021, 15:47:13 UTC |
762e537 | James Rodewig | 01 December 2021, 15:07:41 UTC | [DOCS] Remove source query params from update/delete by query APIs (#81131) (#81212) Removes source-related query parameters from the update by query and delete by query API documentation. These parameters don't return source fields as part of the response. # Conflicts: # rest-api-spec/src/main/resources/rest-api-spec/api/delete_by_query.json | 01 December 2021, 15:07:41 UTC |
04b1b20 | Ignacio Vera | 01 December 2021, 07:00:00 UTC | Spherical mercator transformation should handle properly out of bounds latitudes (#81145) (#81199) Handle properly latitudes close to -90. | 01 December 2021, 07:00:00 UTC |
43ba887 | debadair | 30 November 2021, 22:50:12 UTC | [DOCS] Clarify that pipeline metadata IS required (#81184) | 30 November 2021, 22:50:12 UTC |
49833e3 | Bogdan Pintea | 30 November 2021, 18:25:07 UTC | SQL: Update documentation on Tableau integration (#60890) (#81163) * Update docs on Tableau Desktop integration Update the docs on how to integrate with Tableau Desktop, now using the dedicated connector in conjunction with the JDBC driver. * Add docs for connecting with Tableau Server Add the steps required to connecto to Elasticsearch for Tableau Server. Co-authored-by: James Rodewig <40268737+jrodewig@users.noreply.github.com> (cherry picked from commit fb6baa85aae9581371ffd8d78cbe5df516f828fe) | 30 November 2021, 18:25:07 UTC |
b3708dd | Dimitris Athanasiou | 30 November 2021, 15:47:44 UTC | [7.16] [ML] Parent datafeed actions to the datafeed's persistent task (#81143) (#81153) * [ML] Parent datafeed actions to the datafeed's persistent task (#81143) The vast majority of a datafeed's actions are executed from the data extractor. This includes the heaviest actions which are the searches. This commit passes a `ParentTaskAssigningClient` to `DataExtractorFactory.create` which ensures the client used by any extractor will be setting the corresponding task id: the action task id for preview datafeed and the master operation stage of the start datafeed action, and the persistent task id for the datafeed operations after it has started. * Can't parent on the start datafeed master op as we don't have task | 30 November 2021, 15:47:44 UTC |
deb0d4c | Martijn van Groningen | 30 November 2021, 11:01:49 UTC | [7.16] Add replicated field to get data stream api response. (#81140) * [7.16] Add replicated field to get data stream api response. Backporting #80988 to 7.16 branch. Internally we already kept track of whether a data stream is replicated by CCR. It is part of the `DataStream` class. This just adds it to the xcontent serialization of the get data stream api response class. Relates to elastic/kibana#118899 * adjust to 7.16 reality | 30 November 2021, 11:01:49 UTC |
af5fa3f | Martijn van Groningen | 30 November 2021, 10:49:21 UTC | [7.16] Fix data stream alias validation. (#81040) (#81136) * Fix data stream alias validation. (#81040) In case of restoring a snapshot, it is possible to overwrite an existing data stream with a data stream alias from a snapshot. This change fixes this by improving the generic duplicate name validation. On top of this the lack of data stream alias validation in Metadata.Builder#build() method resulted in cases where data stream aliases could be added for existing index aliases, data streams or indices with the same name. Closes #80972 * adjust to 7.16 reality * Unmute DataStreamsSnapshotsIT#testRestoreDataStreamAliasWithConflictingIndicesAlias() test and fix the test problem, which is that testRestoreDataStreamAliasWithConflictingDataStream() test needs to remove the composable index template that it adds. The base test class doesn't remove any composable index templates and this template interferes with the testRestoreDataStreamAliasWithConflictingIndicesAlias() test. Relates to #81040 | 30 November 2021, 10:49:21 UTC |
b6f463f | Rene Groeschke | 30 November 2021, 09:30:53 UTC | Fix stopping of old elasticsearch cluster (#81059) (#81139) due to not exposing the PID of the underlaying cluster the Fixture Stop task was skipped, leaving running clusters behind after the build finished | 30 November 2021, 09:30:53 UTC |
8ea0930 | Gordon Brown | 29 November 2021, 20:01:31 UTC | Strip blocks from settings for reindex targets (#80887) (#81122) When migrating system features, we copy settings from old indices into the new indices we create before reindexing. However, if we happen to copy a write block, this causes the reindexing to fail. Here, we strip the index block settings before applying settings to new indices. Fixes Co-authored-by: William Brafford <william.brafford@elastic.co> | 29 November 2021, 20:01:31 UTC |
f41a6ed | David Turner | 29 November 2021, 19:44:32 UTC | Clarify `unassigned.reason` docs (#81017) (#81120) Today we indicate that the `unassigned.reason` field in various APIs indicates the reason why a shard is unassigned. This isn't really true, it tells you some information about the event that caused the shard to _become_ unassigned (or which most recently changed its routing table entry while remaining unassigned) but tells you almost nothing about why the shard _is now_ unassigned and how to fix it. That's what the allocation explain API is for. This commit clarifies this point in the docs. Closes #80892 Co-authored-by: James Rodewig <james.rodewig@elastic.co> Co-authored-by: James Rodewig <james.rodewig@elastic.co> | 29 November 2021, 19:44:32 UTC |
ca8cc1f | Lisa Cawley | 29 November 2021, 17:20:31 UTC | [ML] Updates visiblity of validate API (#81061) (#81109) | 29 November 2021, 17:20:31 UTC |
b891026 | James Rodewig | 29 November 2021, 17:03:21 UTC | [DOCS] Update xrefs for snapshot restore docs (#81023) (#81112) Changes: * Removes a leading slash from the restore snapshot API's prerequisites. * Updates several xrefs that point to redirected pages. | 29 November 2021, 17:03:21 UTC |
a9c1279 | David Roberts | 29 November 2021, 16:24:53 UTC | [7.16] [ML] Fix acceptable model snapshot versions in ML deprecation checker (#81105) This is a followup to #81039. The same requirement to tolerate model snapshots back to 6.4.0 that applies to the job opening code also applies to the deprecation checker. Again, we tell the user that 7.0.0 is the model snapshot version we support, but we actually have to support versions going back to 6.4.0 because we didn't update the constant in the C++ in 7.0.0. Additionally, the wording of the ML deprecation messages is very slightly updated. The messages are different in the 7.16 branch, where they were updated by #79387. This wording is copied forward to master, but with the tiny change that "Snapshot" is changed to "Model snapshot" in one place. This should make it clearer for users that we're talking about ML model snapshots and not cluster snapshots (which are completely different things). Another reason to change the wording is that the UI is looking for the pattern /[Mm]odel snapshot/ to decide when to display the "Fix" button for upgrading ML model snapshots - see elastic/kibana#119745. Backport of #81060 | 29 November 2021, 16:24:53 UTC |
3c3d1aa | Dimitris Athanasiou | 29 November 2021, 15:59:21 UTC | [ML] Fix datafeed preview with remote indices (#81099) (#81104) In #77109 a bug was fixed with regard to `date_nanos` time fields and the preview datafeed API. However, that fix introduces a new bug. As we are calling the field caps API to find out whether the time field is `date_nanos`, we are setting the datafeed indices on the request. This may result to erroneous behaviour on local indices and it certainly will result to an error if the datafeed's indices are remote. This commit fixes that problem by setting the datafeed's indices on the field caps request. | 29 November 2021, 15:59:21 UTC |
906eaee | Lisa Cawley | 25 November 2021, 18:05:14 UTC | [DOCS] Clarifies transform ID character limit (#81063) | 25 November 2021, 18:05:14 UTC |
2e750af | Tanguy Leroux | 25 November 2021, 16:22:30 UTC | Fix after restore Lucene.pruneUnreferencedFiles() conditional (#81047) (#81051) In #68821 we introduced a condition to skip the pruning of unreferenced files after the restore of a snapshot for searchable snapshot shards. Sadly I managed to mess this up in a refactoring (#75308) few months after. This commit reintroduces the right conditional which is to NOT prune Lucene files for searchable snapshot shards. | 25 November 2021, 16:22:30 UTC |
85d12e5 | Mark Vieira | 24 November 2021, 21:57:07 UTC | Support additional version schemes in relaxed mode (#81010) (#81021) | 24 November 2021, 21:57:07 UTC |
7f32b68 | James Rodewig | 24 November 2021, 19:47:34 UTC | [DOCS] Fix `type` response values for index recovery API (#81000) (#81015) We updated the `type` response values in https://github.com/elastic/elasticsearch/pull/19516. This updates the docs with the correct values. Closes https://github.com/elastic/elasticsearch/issues/80264 Co-authored-by: David Turner <david.turner@elastic.co> Co-authored-by: David Turner <david.turner@elastic.co> | 24 November 2021, 19:47:34 UTC |
6f8de50 | Christoph Büscher | 24 November 2021, 17:12:51 UTC | Fix flaky DynamicMappingIT testDynamicRuntimeObjectFields (#80999) (#81005) The final part of this test checks that we throw a MapperParsingException when we try to index into a dynamic runtime object that has already been mapped to a different type. Under very rare circumstances this can fail when the mapping update that a previous document index operation has triggered hasn't been completely applied on the shard the second document is targeted at. In this case, indexing the second document can itself trigger a mapping merge operation that can fail with a different exception type (IAE) with a very similar message. In order to simplify the test and make it more robust we can use the same document id for both index requests, making sure we target the same shard group. Closes #80722 | 24 November 2021, 17:12:51 UTC |
988eb19 | Henning Andersen | 24 November 2021, 15:19:44 UTC | Fix flood stage with system indices (#80674) (#81002) System indices do not allow setting changes. This made flood stage handling no longer work for nodes that host a shard of a system index. Now pass in an origin to ensure ES can mark indices located on a node above flood stage as "read-only / allow-delete", regardless of whether it is a system index or not. | 24 November 2021, 15:19:44 UTC |
0876507 | Ievgen Degtiarenko | 24 November 2021, 15:01:44 UTC | clear auto-follow errors on deleting pattern (#80544) | 24 November 2021, 15:01:44 UTC |
6371c2f | Andrei Dan | 24 November 2021, 14:46:36 UTC | Fix WatcherRestartIT (#80986) (#80995) This removes the unnecessary watcher start call in the test to reduce the possibility of flakiness. The call is not needed as the Watcher plugin is active and it will install (and upgrade/remove) the templates regardless if watcher is started. | 24 November 2021, 14:46:36 UTC |
714d7c0 | Joe Gallo | 24 November 2021, 00:01:38 UTC | Use WARNING (rather than CRITICAL) for the empty _tier_preference deprecation issue (#80971) | 24 November 2021, 00:01:38 UTC |
414e838 | Adam Locke | 23 November 2021, 21:39:55 UTC | Fix typo (#80925) (#80974) Co-authored-by: Colin Ng <colin@ngland.net> | 23 November 2021, 21:39:55 UTC |
6776b53 | Mark Vieira | 23 November 2021, 20:53:20 UTC | Increase docker compose timeouts for CI builds | 23 November 2021, 20:54:28 UTC |
dd81287 | James Rodewig | 23 November 2021, 20:02:14 UTC | [DOCS] Relocate `index.mapping.dimension_fields.limit` setting docs (#80964) (#80966) Moves `index.mapping.dimension_fields.limit` so that its co-located with other mapping limit settings. | 23 November 2021, 20:02:14 UTC |
b544968 | Andrei Dan | 23 November 2021, 16:29:44 UTC | Remove obsolete typed legacy index templates (#80949) This removes a few legacy index templates that were superseeded by equivalent component templates or updated index templates. (cherry picked from commit dace65b69aeb0a6c94a9b9878e5cdf0c36b622e2) Signed-off-by: Andrei Dan <andrei.dan@elastic.co> # Conflicts: # x-pack/qa/rolling-upgrade/src/test/java/org/elasticsearch/upgrades/WatcherRestartIT.java | 23 November 2021, 16:29:44 UTC |
b0ed451 | Rory Hunter | 23 November 2021, 11:06:51 UTC | Fix shadowed vars pt6 (#80899) (#80936) Part of #19752. Fix more instances where local variable names were shadowing field names. | 23 November 2021, 11:06:51 UTC |
71c9d9b | Ignacio Vera | 23 November 2021, 09:06:34 UTC | Fix several potential circuit breaker leaks in Aggregators (#79676) (#80929) This commit adds a new CircuitBreaker implementation in the test that throws CircuitBreaker Exceptions randomly. This new circuit breaker helps uncover several places where we might leak if the circuit breaker throws such exception. | 23 November 2021, 09:06:34 UTC |
4fda6e9 | Tim Vernum | 23 November 2021, 00:04:38 UTC | [DOCS] Minor clarifications in LDAP SSL docs (#80897) (#80921) This commit makes a few small changes to the documentation the describes how to configure LDAP with SSL. | 23 November 2021, 00:04:38 UTC |
376c499 | Lisa Cawley | 22 November 2021, 20:18:10 UTC | [DOCS] Fixes query parameters for get buckets API (#80643) (#80917) | 22 November 2021, 20:18:10 UTC |
eaec0d7 | James Rodewig | 22 November 2021, 18:45:34 UTC | [DOCS] Fix typo in gap_policy's default value for serial differencing aggregation (#80893) (#80914) Co-authored-by: James Rodewig <40268737+jrodewig@users.noreply.github.com> Co-authored-by: Simon Stücher <stchr@users.noreply.github.com> | 22 November 2021, 18:45:34 UTC |
1c855f5 | Lisa Cawley | 22 November 2021, 18:09:50 UTC | [DOCS] Adds missing query parameters to ML APIs (#80863) (#80910) | 22 November 2021, 18:09:50 UTC |
41ab64a | Joe Gallo | 22 November 2021, 14:36:33 UTC | Adjust ILMHistoryStore bulk size (#80902) | 22 November 2021, 14:52:20 UTC |