Elasticsearch version 7.12.0
editElasticsearch version 7.12.0
editAlso see Breaking changes in 7.12.
Security updates
edit- A document disclosure flaw was found in the Elasticsearch suggester and profile API when Document and Field Level Security are enabled. The suggester and profile API are normally disabled for an index when document level security is enabled on the index. Certain queries are able to enable the profiler and suggester which could lead to disclosing the existence of documents and fields the attacker should not be able to view. All versions of Elasticsearch before 6.8.15 and 7.11.2 are affected by this flaw. You must upgrade to 6.8.15 or 7.11.2 to obtain the fix. CVE-2021-22137
Known issues
edit-
Frozen tier: (Windows only) The frozen data tier relies on multiple caching mechanisms to speed up access to searchable snapshot files. One of these caches uses sparse files to avoid creating large files on disk when it is not strictly required. A bug prevented files from being created with the right options to enable sparse support on Windows, leading Elasticsearch to create potentially large files that can end up consuming all the disk space.
This issue is fixed in Elasticsearch versions 7.15.2 and later. There is no known workaround for earlier versions. Filesystems that enable sparse files by default are not affected. For more details, see #79371.
-
If autoscaling is enabled for machine learning, the administrator of the cluster
should increase the cluster setting
xpack.ml.max_open_jobs
. This allows autoscaling to run reliably as it relies on assigning jobs only via memory. Havingxpack.ml.max_open_jobs
as a small number may cause autoscaling to behave unexpectedly. - If autoscaling is enabled and job assignment takes an inordinate amount of time, scaling sizes may increase much higher than required. Elasticsearch 7.13.0 contains a fix for this. For more details, see #72509
-
Snapshot and restore: If an index is deleted while the cluster is concurrently taking more than one snapshot then there is a risk that one of the snapshots may never complete and also that some shard data may be lost from the repository, causing future restore operations to fail. To mitigate this problem, set
snapshot.max_concurrent_operations: 1
to prevent concurrent snapshot operations:PUT _cluster/settings { "persistent" : { "snapshot.max_concurrent_operations" : 1 } }
This issue is fixed in Elasticsearch versions 7.13.1 and later. It is not possible to repair a repository once it is affected by this issue, so you must restore the repository from a backup, or clear the repository by executing
DELETE _snapshot/<repository name>/*
, or move to a fresh repository. For more details, see #73456.
Breaking changes
edit- Mapping
- Query Languages
- SQL
Breaking Java changes
edit- Features/Java High Level REST Client
-
- Move find file structure finder in REST High Level client to its new endpoint and plugin #67290
Deprecations
edit- Distributed
- Infra/Core
- Geo
- Machine Learning
- Security
New features
edit- Authorization
- EQL
- Features/ILM+SLM
- Features/Ingest
- Query Languages
- SQL
- Search
- Snapshot/Restore
-
- Add partial searchable snapshot support for a frozen tier #68509
Enhancements
edit- Aggregations
- Allocation
- Authentication
-
- Add grant-api-key to HLRC #68190
- Authorization
-
- Add more context to cluster access denied messages #66900 (issues: #42166, #60357)
- Add more context to index access denied errors #60357 (issue: #42166)
- Allow Kibana to create Fleet Server indices #68152
- Phase 2 support for operator privileges: Cluster settings #66684
-
view_index_metadata
andmanage
privs now grant access to field capabilities #67392 (issue: #66867)
- Cluster Coordination
- EQL
- Engine
- Features/ILM+SLM
- Features/Indices APIs
- Features/Ingest
- Features/Java High Level REST Client
- Features/Java Low Level REST Client
- Features/Stats
- Geo
- Highlighting
- Infra/Core
- Infra/Logging
- Infra/Plugins
- Infra/Scripting
- Infra/Settings
- License
-
- Remove grace period from license expiration check #67316
- Machine Learning
-
- Add new multi custom processor for data frame analytics and model inference #67362
- Allow data frame analytics memory estimation on non-ML nodes #68146 (issue: #67840)
- Automatic management for machine learning system indices #68044
- Expand regression and classification hyperparameters #67950
- Improve resuming a data frame analytics job stopped during inference #67623
- Move find file structure to a new API endpoint #67123 (issue: #67001)
- Preserve inference progress when starting data frame analytics #68222
- Write notification messages indicating required capacity when a job is not assigned to a node #67181
- Fix edge case which could cause spurious anomalies early in the learning process if the time series has non-diurnal seasonality #1634
- Compute importance of hyperparameters optimized in the fine parameter tuning step #1627
- Early stopping for the fine parameter tuning step of classification and regression model training #1676
- Correct upgrade for pre-6.3 state for lat_long anomaly anomaly detectors #1681
- Per tree feature bag to speed up training of regression and classification models and improve scalability for large numbers of features #1733
- Mapping
-
- Shortcut to avoid fmod #66909
- Packaging
-
- Tweaks to the Iron Bank docker context #66942
- Performance
-
- Add benchmark racing scripts #68369
- Query Languages
- Ranking
- Rollup
-
- Remove the rollup thread pool #65958
- SQL
- Search
-
- Add grok and dissect methods to runtime fields #68088 (issue: #67825)
- Allow deletion of async searches with the manage privilege #67965
-
Avoid duplicate serialization for
TermsQueryBuilder
#67223 (issue: #67132) - Early detection of circuit breaker exception in the coordinating node #67431 (issue: #62884)
- Limit the depth of nested bool queries #66204 (issue: #55303)
-
Remove unnecessary optimizations for
TermsSetQueryBuilder
#67637 (issue: #67223) - Retry point in time on other copy when possible #66713 (issue: #61062)
- Slightly speed up scripts #68478
- Use a mapping snapshot for fetching nested docs #66877 (issue: #66295)
- Security
- Snapshot/Restore
-
-
Add ClusterUUID to
RepositoryData
#68002 - Allow searchable snapshots to work on source-only repositories #67975 (issue: #67936)
- Avoid async cache-size fetch on partial shards #68644
- Forbid removing write block from searchable snapshot index #68036
- Ignore disk watermarks on partial shards #68673
- Improve error message on incompatible repo format #67776
- Introduce repository UUIDs #67829 (issue: #66431)
- Introduce repository test kit/analyser #67247
- Permit rename of repository of searchable snapshot #67968 (issue: #66431)
- Reject remounting snapshot of a searchable snapshot #68816 (issue: #68792)
-
Add ClusterUUID to
- Task Management
- Transform
-
- Add support for search-time runtime fields #67643 (issue: #65147)
- Automatic management for transform system indices #68590
- Implement retention policy to delete data from a transform #67832 (issue: #67916)
- Refactor cat transform to show more useful information #68232 (issues: #66367, #66718)
- Report last search time in transform stats #66718 (issues: #66367, #66410, #66718)
Bug fixes
edit- Aggregations
- Engine
-
- Fail force-merges on read-only engines #64756
- Features/Features
- Features/ILM+SLM
- Features/Java Low Level REST Client
- Features/Monitoring
- Features/Ingest
-
- Fix ingest java week based year defaulting #65717
- Geo
- Infra/Core
- Infra/Scripting
- Machine Learning
-
-
Make
find_structure
validatelines_to_sample
at least 2 #68199 (issue: #68128) -
Return 400 status for invalid charset in
find_structure
#68201 (issue: #68130) -
Return status 400 for
grok_pattern
errors infind_structure
#68205 (issue: #68132) - Fix a source of instability in time series modeling for anomaly detection. This has been observed to cause spurious anomalies for a partition which no longer receives any data #1675
- Ensure that we stop modeling seasonality for data which flatlines. This is important for count and sum detectors which treat empty buckets as zero. Before this change, we could sometimes detect spurious anomalies in realtime detection after a partition no longer received any data #1654
-
Make
- SQL
- Search
-
-
Handle ignored fields directly in
SourceValueFetcher
#68738
-
Handle ignored fields directly in
- Snapshot/Restore