Elasticsearch version 8.0.0

Elasticsearch version 8.0.0

The following list are changes in 8.0.0 as compared to 7.17.0, and combines release notes from the 8.0.0-alpha1, -alpha2, -beta1, -rc1 and -rc2 releases.

Also see Breaking changes in 8.0.

Known issues

  • If you installed Elasticsearch from an archive on an aarch64 platform like Linux ARM or macOS M1, the elastic user password and Kibana enrollment token are not generated automatically when starting your node for the first time.

    After the node starts, generate the elastic password with the bin/elasticsearch-reset-password tool:

    1. bin/elasticsearch-reset-password -u elastic

    Then, create an enrollment token for Kibana with the bin/elasticsearch-create-enrollment-token tool:

    1. bin/elasticsearch-create-enrollment-token -s kibana

  • Parsing a request when the last element in an array is filtered out (for instance using _source_includes) fails. This is due to a bug in Jackson parser. Fixed in Elasticsearch 8.6.1 (#91456)

Breaking changes

Aggregations

  • Percentiles aggregation: disallow specifying same percentile values twice #52257 (issue: #51871)
  • Remove adjacency matrix setting #46327 (issues: #46257, #46324)
  • Remove MovingAverage pipeline aggregation #39328
  • Remove deprecated _time and _term sort orders #39450
  • Remove deprecated date histo interval #75000

Allocation

Analysis

Authentication

  • Always add file and native realms unless explicitly disabled #69096 (issue: #50892)
  • Do not set a NameID format in Policy by default #44090 (issue: #40353)
  • Make order setting mandatory for Realm config #51195 (issue: #37614)

CCR

  • Avoid auto following leader system indices in CCR #72815 (issue: #67686)

Cluster Coordination

Distributed

Engine

  • Force merge should reject requests with only_expunge_deletes and max_num_segments set #44761 (issue: #43102)
  • Remove per-type indexing stats #47203 (issue: #41059)
  • Remove translog retention settings #51697 (issue: #50775)

Features/CAT APIs

  • Remove the deprecated local parameter for _cat/indices #64868 (issue: #62198)
  • Remove the deprecated local parameter for _cat/shards #64867 (issue: #62197)

Features/Features

  • Remove deprecated ._tier allocation filtering settings #73074 (issue: #72835)

Features/ILM+SLM

  • Add lower bound on poll_interval #39593 (issue: #39163)
  • Make the ILM freeze action a no-op #77158 (issue: #70192)
  • Always enforce default tier preference #79751 (issue: #76147)
  • Validate that snapshot repository exists for ILM policies at creation/update time #78468 (issues: #72957, #77657)
  • Default cluster.routing.allocation.enforce_default_tier_preference to true #79275 (issues: #76147, #79210)

Features/Indices APIs

  • Change prefer_v2_templates parameter to default to true #55489 (issues: #53101, #55411)
  • Remove deprecated _upgrade API #64732 (issue: #21337)
  • Remove local parameter for get field mapping request #55100 (issue: #55099)
  • Remove include_type_name parameter from REST layer #48632 (issue: #41059)
  • Remove the template field in index templates #49460 (issue: #21009)
  • Remove endpoint for freezing indices #78918 (issues: #70192, #77273)

Features/Watcher

  • Move watcher history to data stream #64252

Geo

  • Disallow creating geo_shape mappings with deprecated parameters #70850 (issue: #32039)
  • Remove bounding box query type parameter #74536

Infra/Circuit Breakers

  • Fixed synchronizing inflight breaker with internal variable #40878

Infra/Core

  • Limit processors by available processors #44894 (issue: #44889)
  • Remove nodes/0 folder prefix from data path #42489
  • Remove bootstrap.system_call_filter setting #72848
  • Remove fixed_auto_queue_size threadpool type #52280
  • Remove node.max_local_storage_nodes #42428 (issue: #42426)
  • Remove camel case named date/time formats #60044
  • Remove legacy role settings #71163 (issues: #54998, #66409, #71143)
  • Remove processors setting #45905 (issue: #45855)
  • Remove the local parameter of /_cat/nodes #50594 (issues: #50088, #50499)
  • Remove the listener thread pool #53314 (issue: #53049)
  • Remove the node local storage setting #54381 (issue: #54374)
  • Remove the pidfile setting #45940 (issue: #45938)
  • Removes week_year date format #63384 (issue: #60707)
  • System indices treated as restricted indices #74212 (issue: #69298)
  • Remove Joda dependency #79007
  • Remove Joda support from date formatters #78990
  • All system indices are hidden indices #79512

Infra/Logging

Infra/Plugins

  • Remove deprecated basic license feature enablement settings #56211 (issue: #54745)

Infra/REST API

  • Remove content type required setting #61043
  • Remove deprecated endpoints containing _xpack #48170 (issue: #35958)
  • Remove deprecated endpoints of hot threads API #55109 (issue: #52640)
  • Allow parsing Content-Type and Accept headers with version #61427

Infra/Resiliency

Infra/Scripting

Infra/Settings

  • Change default value of action.destructive_requires_name to true #66908 (issue: #61074)
  • Forbid settings without a namespace #45947 (issues: #45905, #45940)

Ingest

  • Remove default maxmind GeoIP databases from distribution #78362 (issue: #68920)

License

  • Set xpack.security.enabled to true for all licenses #72300
  • Enforce license expiration #79671

Machine Learning

  • Remove deprecated _xpack endpoints #59870 (issues: #35958, #48170)
  • Remove the ability to update datafeed’s job_id #44752 (issue: #44616)
  • Remove allow_no_datafeeds and allow_no_jobs parameters from APIs #80048 (issue: #60732)

Mapping

  • Remove boost mapping parameter #62639 (issue: #62623)
  • Remove support for chained multi-fields #42333 (issues: #41267, #41926)
  • Remove support for string in unmapped_type #45675
  • Removes typed URLs from mapping APIs #41676

Network

Packaging

Recovery

  • Remove dangling index auto import functionality #59698 (issue: #48366)

Reindex

Rollup

  • RollupStart endpoint should return OK if job already started #41502 (issues: #35928, #39845)

Search

  • Decouple shard allocation awareness from search and get requests #45735 (issue: #43453)
  • Fix range query on date fields for number inputs #63692 (issue: #63680)
  • Make fuzziness reject illegal values earlier #33511
  • Make remote cluster resolution stricter #40419 (issue: #37863)
  • Parse empty first line in msearch request body as action metadata #41011 (issue: #39841)
  • Remove CommonTermsQuery and cutoff_frequency param #42654 (issue: #37096)
  • Remove type query #47207 (issue: #41059)
  • Remove use_field_mapping format option for docvalue fields #55622
  • Remove deprecated SimpleQueryStringBuilder parameters #57200
  • Remove deprecated search.remote settings #42381 (issues: #33413, #38556)
  • Remove deprecated sort options: nested_path and nested_filter #42809 (issue: #27098)
  • Remove deprecated vector functions #48725 (issue: #48604)
  • Remove support for _type in searches #68564 (issues: #41059, #68311)
  • Remove support for sparse vectors #48781 (issue: #48368)
  • Remove the object format for indices_boost #55078
  • Removes type from TermVectors APIs #42198 (issue: #41059)
  • Removes typed endpoint from search and related APIs #41640
  • Set max allowed size for stored async response #74455 (issue: #67594)
  • indices.query.bool.max_clause_count now limits all query clauses #75297

Security

  • Remove obsolete security settings #40496
  • Remove support of creating CA on the fly when generating certificates #65590 (issue: #61884)
  • Remove the id field from the InvalidateApiKey API #66671 (issue: #66317)
  • Remove the migrate tool #42174
  • Compress audit logs #64472 (issue: #63843)
  • Remove insecure settings #46147 (issue: #45947)
  • Remove kibana_dashboard_only_user reserved role #76507

Snapshot/Restore

  • Blob store compress default to true #40033
  • Get snapshots support for multiple repositories #42090 (issue: #41210)
  • Remove repository stats API #62309 (issue: #62297)
  • Remove frozen cache setting leniency #71013 (issue: #70341)
  • Adjust snapshot index resolution behavior to be more intuitive #79670 (issue: #78320)

TLS

  • Reject misconfigured/ambiguous SSL server config #45892
  • Remove support for configurable PKCS#11 keystores #75404
  • Remove the client transport profile filter #43236

Breaking Java changes

Authentication

  • Mandate x-pack REST handler installed #71061 (issue: #70523)

CCR

CRUD

  • Remove types from BulkRequest #46983 (issue: #41059)
  • Remove Client.prepareIndex(index, type, id) method #48443

Client

  • Remove SecurityClient from x-pack #42471

Features/ILM+SLM

Features/Monitoring

  • Remove MonitoringClient from x-pack #42770

Features/Watcher

  • Remove WatcherClient from x-pack #42815

Infra/Core

  • Remove XPackClient from x-pack #42729
  • Remove the transport client #42538
  • Remove transport client from x-pack #42202

Infra/REST API

  • Copy HTTP headers to ThreadContext strictly #45945

Machine Learning

  • Remove the MachineLearningClient #43108

Mapping

  • Remove type filter from GetMappings API #47364 (issue: #41059)
  • Remove type parameter from PutMappingRequest.buildFromSimplifiedDef() #50844 (issue: #41059)
  • Remove unused parameter from MetadataFieldMapper.TypeParser#getDefault() #51219
  • Remove type parameter from CIR.mapping(type, object...) #50739 (issue: #41059)

Search

  • Removes types from SearchRequest and QueryShardContext #42112

Snapshot/Restore

  • Remove deprecated repository methods #42359 (issue: #42213)

Deprecations

Authentication

  • Deprecate setup-passwords tool #76902

CRUD

  • Remove indices_segments verbose parameter #78451 (issue: #75955)

Engine

  • Deprecate setting max_merge_at_once_explicit #80574

Machine Learning

  • Deprecate estimated_heap_memory_usage_bytes and replace with model_size_bytes #80554

Monitoring

  • Add deprecation info API entries for deprecated monitoring settings #78799
  • Automatically install monitoring templates at plugin initialization #78350
  • Remove Monitoring ingest pipelines #77459 (issue: #50770)

Search

  • Configure IndexSearcher.maxClauseCount() based on node characteristics #81525 (issue: #46433)

Transform

New features

Security

Snapshot/Restore

  • Support IAM roles for Kubernetes service accounts #81255 (issue: #52625)

Watcher

  • Use startsWith rather than exact matches for Watcher history template names #82396

Enhancements

Analysis

  • Move reload_analyzers endpoint to x-pack #43559

Authentication

  • Reset elastic password CLI tool #74892 (issues: #70113, #74890)
  • Autogenerate and print elastic password on startup #77291
  • Enroll Kibana API uses Service Accounts #76370
  • Add reset-kibana-system-user tool #77322
  • New CLI tool to reset password for built-in users #79709
  • Auto-configure the elastic user password #78306

Authorization

  • Granting kibana_system reserved role access to “all” privileges to .internal.preview.alerts* index #80889 (issues: #76624, #80746, #116374)
  • Granting kibana_system reserved role access to “all” privileges to .preview.alerts* index #80746
  • Granting editor and viewer roles access to alerts-as-data indices #81285

Cluster Coordination

Data streams

  • Data stream support read and write with custom routing and partition size #74394 (issue: #74390)

EQL

  • Add option for returning results from the tail of the stream #64869 (issue: #58646)
  • Introduce case insensitive variant in~ #68176 (issue: #68172)
  • Optimize redundant toString #71070 (issue: #70681)

Engine

  • Always use soft-deletes in InternalEngine #50415
  • Remove translog retention policy #51417 (issue: #50775)

Features/CAT APIs

  • Remove size and add time params to _cat/threadpool #55736 (issue: #54478)

Features/ILM+SLM

  • Allow for setting the total shards per node in the Allocate ILM action #76794 (issue: #76775)
  • Inject migrate action regardless of allocate action #79090 (issue: #76147)
  • Make unchanged ILM policy updates into noop #82240 (issue: #82065)
  • Avoid unnecessary LifecycleExecutionState recalculation #81558 (issues: #77466, #79692)

Features/Indices APIs

Features/Stats

Features/Watcher

  • Remove Watcher history clean up from monitoring #67154

Infra/Core

  • Remove aliases exist action #43430
  • Remove indices exists action #43164
  • Remove types exists action #43344
  • Retain reference to stdout for exceptional cases #77460
  • Check whether stdout is a real console #79882
  • Share int, long, float, double, and byte pages #75053
  • Revert “Deprecate resolution loss on date field (#78921)” #79914 (issue: #78921)
  • Add two missing entries to the deprecation information API #80290 (issue: #80233)
  • Prevent upgrades to 8.0 without first upgrading to the last 7.x release #82321 (issue: #81865)

Infra/Logging

  • Make Elasticsearch JSON logs ECS compliant #47105 (issue: #46119)

Infra/REST API

  • Allow for field declaration for future compatible versions #69774 (issue: #51816)
  • Introduce stability description to the REST API specification #38413
  • Parsing: Validate that fields are not registered twice #70243
  • Support response content-type with versioned media type #65500 (issue: #51816)
  • [REST API Compatibility] Typed endpoints for index and get APIs #69131 (issue: #54160)
  • [REST API Compatibility] Typed endpoints for put and get mapping and get field mappings #71721 (issues: #51816, #54160)
  • [REST API Compatibility] Allow copy_settings flag for resize operations #75184 (issues: #38514, #51816)
  • [REST API Compatibility] Allow for type in geo shape query #74553 (issues: #51816, #54160)
  • [REST API Compatibility] Always return adjust_pure_negative value #75182 (issues: #49543, #51816)
  • [REST API Compatibility] Clean up x-pack/plugin rest compat tests #74701 (issue: #51816)
  • [REST API Compatibility] Do not return _doc for empty mappings in template #75448 (issues: #51816, #54160, #70966, #74544)
  • [REST API Compatibility] Dummy REST action for indices.upgrade API #75136 (issue: #51816)
  • [REST API Compatibility] REST Terms vector typed response #73117
  • [REST API Compatibility] Rename BulkItemResponse.Failure type field #74937 (issue: #51816)
  • [REST API Compatibility] Type metadata for docs used in simulate request #74222 (issues: #51816, #54160)
  • [REST API Compatibility] Typed TermLookups #74544 (issues: #46943, #51816, #54160)
  • [REST API Compatibility] Typed and x-pack graph explore API #74185 (issues: #46935, #51816, #54160)
  • [REST API Compatibility] Typed endpoint for bulk API #73571 (issue: #51816)
  • [REST API Compatibility] Typed endpoint for multi-get API #73878 (issue: #51816)
  • [REST API Compatibility] Typed endpoints for RestUpdateAction and RestDeleteAction #73115 (issues: #51816, #54160)
  • [REST API Compatibility] Typed endpoints for get_source API #73957 (issues: #46587, #46931, #51816)
  • [REST API Compatibility] Typed endpoints for explain API #73901 (issue: #51816)
  • [REST API Compatibility] Typed endpoints for search _count API #73958 (issues: #42112, #51816)
  • [REST API Compatibility] Typed indexing stats #74181 (issues: #47203, #51816, #54160)
  • [REST API Compatibility] Types for percolate query API #74698 (issues: #46985, #51816, #54160, #74689)
  • [REST API Compatibility] Validate query typed API #74171 (issues: #46927, #51816, #54160)
  • [REST API Compatibility] Voting config exclusion exception message #75406 (issues: #51816, #55291)
  • [REST API Compatibility] MoreLikeThisQuery with types #75123 (issues: #42198, #51816, #54160)
  • [REST API Compatibility] Update and delete by query using size field #69606
  • [REST API Compatibility] Indicies boost in object format #74422 (issues: #51816, #55078)
  • [REST API Compatibility] Typed endpoints for search and related endpoints #72155 (issues: #51816, #54160)
  • [REST API Compatibility] Allow to use size -1 #75342 (issues: #51816, #69548, #70209)
  • [REST API Compatibility] Ignore use_field_mapping option for docvalue #74435 (issue: #55622)
  • [REST API Compatibility] _time and _term sort orders #74919 (issues: #39450, #51816)
  • [REST API Compatability] template parameter and field on PUT index template #71238 (issues: #49460, #51816, #68905)
  • [REST API Compatibility] Make query registration easier #75722 (issue: #51816)
  • [REST API Compatibility] Typed query #75453 (issues: #47207, #51816, #54160)
  • [REST API Compatibility] Deprecate the use of synced flush #75372 (issues: #50882, #51816)
  • [REST API Compatibility] Licence accept_enterprise and response changes #75479 (issues: #50067, #50735, #51816, #58217)

Infra/Scripting

Infra/Settings

  • Fixed inconsistent Setting.exist() #46603 (issue: #41830)
  • Remove index.optimize_auto_generated_id setting (#27583) #27600 (issue: #27583)
  • Implement setting deduplication via string interning #80493 (issues: #77466, #78892)

Ingest

  • Add support for _meta field to ingest pipelines #76381
  • Remove binary field after attachment processor execution #79172
  • Improving cache lookup to reduce recomputing / searches #77259
  • Extract more standard metadata from binary files #78754 (issue: #22339)

License

  • Add deprecated accept_enterprise param to /_xpack #58220 (issue: #58217)
  • Support accept_enterprise param in get license API #50067 (issue: #49474)
  • Enforce Transport TLS check on all licenses #79602 (issue: #75292)

Machine Learning

  • The Windows build platform for the machine learning C++ code now uses Visual Studio 2019 #1352
  • The macOS build platform for the machine learning C++ code is now Mojave running Xcode 11.3.1, or Ubuntu 20.04 running clang 8 for cross compilation #1429
  • Add a new application for evaluating PyTorch models. The app depends on LibTorch - the C++ front end to PyTorch - and performs inference on models stored in the TorchScript format #1902
  • Adding new PUT trained model vocabulary endpoint #77387
  • Creating new PUT model definition part API #76987
  • Add inference time configuration overrides #78441 (issue: #77799)
  • Optimize source extraction for categorize_text aggregation #79099
  • The Linux build platform for the machine learning C++ code is now CentOS 7 running gcc 10.3. #2028
  • Make ML indices hidden when the node becomes master #77416 (issue: #53674)
  • Add deployment_stats to trained model stats #80531
  • The setting use_auto_machine_memory_percent now defaults max_model_memory_limit #80532 (issue: #80415)

Mapping

  • Sparse vector to throw exception consistently #62646
  • Add support for configuring HNSW parameters #79193 (issue: #78473)
  • Extend dense_vector to support indexing vectors #78491 (issue: #78473)

Monitoring

  • Add previously removed Monitoring settings back for 8.0 #78784
  • Change Monitoring plugin cluster alerts to not install by default #79657
  • Adding default templates for Metricbeat ECS data #81744

Network

  • Enable LZ4 transport compression by default #76326 (issue: #73497)
  • Improve slow inbound handling to include response type #80425

Packaging

  • Make the Docker build more re-usable in Cloud #50277 (issues: #46166, #49926)
  • Update docker-compose.yml to fix bootstrap check error #47650
  • Allow total memory to be overridden #78750 (issue: #65905)
  • Convert repository plugins to modules #81870 (issue: #81652)

Recovery

  • Use Lucene index in peer recovery and resync #51189 (issue: #50775)
  • Fix PendingReplicationActions submitting lots of NOOP tasks to GENERIC #82092 (issues: #77466, #79837)

Reindex

Rollup

  • Adds support for date_nanos in Rollup Metric and DateHistogram Configs #59349 (issue: #44505)

SQL

  • Add text formatting support for multivalue #68606
  • Add xDBC and CLI support. QA CSV specs #68966
  • Export array values through result sets #69512
  • Improve alias resolution in sub-queries #67216 (issue: #56713)
  • Improve the optimization of null conditionals #71192
  • Push WHERE clause inside subqueries #71362
  • Use Java String methods for LTRIM/RTRIM #57594
  • QL: Make canonical form take into account children #71266
  • QL: Polish optimizer expression rule declaration #71396
  • QL: Propagate nullability constraints across conjunctions #71187 (issue: #70683)

Search

  • Completely disallow setting negative size in search #70209 (issue: #69548)
  • Make 0 as invalid value for min_children in has_child query #41347
  • Return error when remote indices are locally resolved #74556 (issue: #26247)
  • [REST API Compatibility] Nested path and filter sort options #76022 (issues: #42809, #51816)
  • [REST API Compatibility] CommonTermsQuery and cutoff_frequency parameter #75896 (issues: #42654, #51816)
  • [REST API Compatibility] Allow first empty line for _msearch #75886 (issues: #41011, #51816)
  • Node level can match action #78765
  • TSDB: Add time series information to field caps #78790 (issue: #74660)
  • Add new kNN search endpoint #79013 (issue: #78473)
  • Disallow kNN searches on nested vector fields #79403 (issue: #78473)
  • Ensure kNN search respects authorization #79693 (issue: #78473)
  • Load kNN vectors format with mmapfs #78724 (issue: #78473)
  • Support cosine similarity in kNN search #79500
  • Node level can match action #78765
  • Check nested fields earlier in kNN search #80516 (issue: #78473)

Security

  • Add a tool for creating enrollment tokens #74890
  • Add the Enroll Kibana API #72207
  • Change default hashing algorithm for FIPS 140 #55544
  • Create enrollment token #73573 (issues: #71438, #72129)
  • Enroll node API #72129
  • Configure security for the initial node CLI #74868
  • Generate and store password hash for elastic user #76276 (issue: #75310)
  • Set elastic password and generate enrollment token #75816 (issue: #75310)
  • Add elasticsearch-enroll-node tool #77292
  • Default hasher to PBKDF2_STRETCH on FIPS mode #76274
  • Add v7 restCompat for invalidating API key with the id field #78664 (issue: #66671)
  • Print enrollment token on startup #78293
  • Startup check for security implicit behavior change #76879
  • CLI tool to reconfigure nodes to enroll #79690 (issue: #7718)
  • Security auto-configuration for packaged installations #75144 (issue: #78306)

Snapshot/Restore

  • Introduce searchable snapshots index setting for cascade deletion of snapshots #74977
  • Unify blob store compress setting #39346 (issue: #39073)
  • Add recovery state tracking for searchable snapshots #60505
  • Allow listing older repositories #78244
  • Optimize SLM Policy Queries #79341 (issue: #79321)

TLS

  • Add ChaCha20 TLS ciphers on Java 12+ #42155
  • Add support for KeyStore filters to ssl-config #75407
  • Update TLS ciphers and protocols for JDK 11 #41808 (issues: #38646, #41385)

Transform

  • Prevent old beta transforms from starting #79712

TSDB

  • Automatically add timestamp mapper #79136
  • Create a coordinating node level reader for tsdb #79197
  • Fix TSDB shrink test in multi-version cluster #79940 (issue: #79936)
  • Do not allow shadowing metrics or dimensions #79757

Bug fixes

Aggregations

  • Fix BWC issues for x_pack/usage #55181 (issue: #54847)
  • Fix DoubleBounds null serialization #59475
  • Fix TopHitsAggregationBuilder adding duplicate _score sort clauses #42179 (issue: #42154)
  • Fix t_test usage stats #54753 (issue: #54744)
  • Throw exception if legacy interval cannot be parsed in DateIntervalWrapper #41972 (issue: #41970)

Autoscaling

  • Autoscaling use adjusted total memory #80528 (issue: #78750)

CCR

Cluster Coordination

  • Apply cluster states in system context #53785 (issue: #53751)

Data streams

  • Prohibit restoring a data stream alias with a conflicting write data stream #81217 (issue: #80976)

Distributed

  • Introduce ?wait_for_active_shards=index-setting #67158 (issue: #66419)
  • Fixes to task result index mapping #50359 (issue: #50248)

Features/CAT APIs

  • Fix cat recovery display of bytes fields #40379 (issue: #40335)

Features/ILM+SLM

  • Ensuring that the ShrinkAction does not hang if total shards per node is too low #76732 (issue: #44070)
  • Less verbose serialization of snapshot failure in SLM metadata #80942 (issue: #77466)

Features/Indices APIs

  • Fix ComposableIndexTemplate equals when composed_of is null #80864

Features/Java High Level REST Client

Geo

  • Preprocess polygon rings before processing it for decomposition #59501 (issues: #54441, #59386)

Infra/Core

  • Add searchable snapshot cache folder to NodeEnvironment #66297 (issue: #65725)
  • CLI tools: Write errors to stderr instead of stdout #45586 (issue: #43260)
  • Precompute ParsedMediaType for XContentType #67409
  • Prevent stack overflow in rounding #80450

Infra/Logging

  • Fix NPE when logging null values in JSON #53715 (issue: #46702)
  • Fix stats in slow logs to be a escaped JSON #44642
  • Populate data stream fields when xOpaqueId not provided #62156

Infra/REST API

Infra/Scripting

  • Change compound assignment structure to support string concatenation #61825
  • Fixes casting in constant folding #61508
  • Several minor Painless fixes #61594
  • Fix duplicated allow lists upon script engine creation #82820 (issue: #82778)

Infra/Settings

  • Stricter UpdateSettingsRequest parsing on the REST layer #79227 (issue: #29268)
  • Set Auto expand replica on deprecation log data stream #79226 (issue: #78991)

Ingest

  • Adjust default geoip logging to be less verbose #81404 (issue: #81356)

Machine Learning

  • Add timeout parameter for delete trained models API #79739 (issue: #77070)
  • Tone down ML unassigned job notifications #79578 (issue: #79270)
  • Use a new annotations index for future annotations #79006 (issue: #78439)
  • Set model state compatibility version to 8.0.0 #2139
  • Check that total_definition_length is consistent before starting a deployment #80553
  • Fail inference processor more consistently on certain error types #81475
  • Optimize the job stats call to do fewer searches #82362 (issue: #82255)

Mapping

  • Remove assertions that mappings have one top-level key #58779 (issue: #58521)

Packaging

  • Suppress illegal access in plugin install #41620 (issue: #41478)

Recovery

  • Make shard started response handling only return after the cluster state update completes #82790 (issue: #81628)

SQL

  • Introduce dedicated node for HAVING declaration #71279 (issue: #69758)
  • Make RestSqlQueryAction thread-safe #69901

Search

  • Check for negative from values in search request body #54953 (issue: #54897)
  • Fix VectorsFeatureSetUsage serialization in BWC mode #55399 (issue: #55378)
  • Handle total hits equal to track_total_hits #37907 (issue: #37897)
  • Improve error msg for CCS request on node without remote cluster role #60351 (issue: #59683)
  • Remove unsafe assertion in wildcard field #78966

Security

  • Allow access to restricted system indices for reserved system roles #76845

Snapshot/Restore

  • Fix GET /_snapshot/_all/_all if there are no repos #43558 (issue: #43547)
  • Don’t fill stack traces in SnapshotShardFailure #80009 (issue: #79718)
  • Remove custom metadata if there is nothing to restore #81373 (issues: #81247, #82019)

Regressions

Search

  • Disable numeric sort optimization conditionally #78103

Upgrades

Authentication

  • Upgrade to UnboundID LDAP SDK v6.0.2 #79332

Infra/Logging

  • Upgrade ECS logging layout to latest version #80500

Search

Security

Snapshot/Restore

  • Upgrade repository-hdfs plugin to Hadoop 3 #76897