elasticsearch/docs/reference/release-notes/8.15.1.asciidoc

96 lines
4.9 KiB
Text

[[release-notes-8.15.1]]
== {es} version 8.15.1
Also see <<breaking-changes-8.15,Breaking changes in 8.15>>.
[[known-issues-8.15.1]]
[float]
=== Known issues
* Elasticsearch will not start if custom role mappings are configured using the
`xpack.security.authc.realms.*.files.role_mapping` configuration option. As a workaround, custom role mappings
can be configured using the https://www.elastic.co/guide/en/elasticsearch/reference/current/security-api-put-role-mapping.html[REST API] (issue: {es-issue}112503[#112503])
* ES|QL queries can lead to node crashes due to Out Of Memory errors when:
** Multiple indices match the query pattern
** These indices have many conflicting field mappings
** Many of those fields are included in the request
These issues deplete heap memory, increasing the likelihood of OOM errors. (issue: {es-issue}111964[#111964], {es-issue}111358[#111358]).
In Kibana, you might indirectly execute these queries when using Discover, or adding a Field Statistics panel to a dashboard.
+
To work around this issue, you have a number of options:
** Downgrade to an earlier version
** Upgrade to 8.15.2 upon release
** Follow the instructions to
<<esql-kibana-enable,disable ES|QL queries in {kib}>>
** Change the default data view in Discover to a smaller set of indices and/or one with fewer mapping conflicts.
* Index Stats, Node Stats and Cluster Stats API can return a null pointer exception if an index contains a `dense_vector` field
but there is an index segment that does not contain any documents with a dense vector field ({es-pull}112720[#112720]). Workarounds:
** If the affected index already contains documents with a dense vector field, force merge the index to a single segment.
** If the affected index does not already contain documents with a dense vector field, index a document with a dense vector field
and then force merge to a single segment.
** If the affected index's `dense_vector` fields are unused, reindex without the `dense_vector` fields.
[[bug-8.15.1]]
[float]
=== Bug fixes
Aggregations::
* Revert "Avoid bucket copies in Aggs" {es-pull}111758[#111758] (issue: {es-issue}111679[#111679])
Authorization::
* Fix DLS over Runtime Fields {es-pull}112260[#112260] (issue: {es-issue}111637[#111637])
ES|QL::
* Avoid losing error message in failure collector {es-pull}111983[#111983] (issue: {es-issue}111894[#111894])
* Avoid wrapping rejection exception in exchange {es-pull}112178[#112178] (issue: {es-issue}112106[#112106])
* ESQL: Fix for overzealous validation in case of invalid mapped fields {es-pull}111475[#111475] (issue: {es-issue}111452[#111452])
Geo::
* Add maximum nested depth check to WKT parser {es-pull}111843[#111843]
* Always check `crsType` when folding spatial functions {es-pull}112090[#112090] (issue: {es-issue}112089[#112089])
* Fix NPE when executing doc value queries over shape geometries with empty segments {es-pull}112139[#112139]
Indices APIs::
* Fix template alias parsing livelock {es-pull}112217[#112217]
Infra/Core::
* Fix windows memory locking {es-pull}111866[#111866] (issue: {es-issue}111847[#111847])
Ingest Node::
* Fixing incorrect bulk request took time {es-pull}111863[#111863] (issue: {es-issue}111854[#111854])
* Improve performance of grok pattern cycle detection {es-pull}111947[#111947]
Logs::
* Merge multiple ignored source entires for the same field {es-pull}111994[#111994] (issue: {es-issue}111694[#111694])
Machine Learning::
* [Inference API] Move Delete inference checks to threadpool worker {es-pull}111646[#111646]
Mapping::
* Check for valid `parentDoc` before retrieving its previous {es-pull}112005[#112005] (issue: {es-issue}111990[#111990])
* Fix calculation of parent offset for ignored source in some cases {es-pull}112046[#112046]
* Fix synthetic source for empty nested objects {es-pull}111943[#111943] (issue: {es-issue}111811[#111811])
* No error when `store_array_source` is used without synthetic source {es-pull}111966[#111966]
* Prevent synthetic field loaders accessing stored fields from using stale data {es-pull}112173[#112173] (issue: {es-issue}112156[#112156])
Ranking::
* Properly handle filters on `TextSimilarityRank` retriever {es-pull}111673[#111673]
Relevance::
* Semantic reranking should fail whenever inference ID does not exist {es-pull}112038[#112038] (issue: {es-issue}111934[#111934])
* [Bugfix] Add `accessDeclaredMembers` permission to allow search application templates to parse floats {es-pull}111285[#111285]
Search::
* Explain Function Score Query {es-pull}111807[#111807]
Security::
* Fix "unexpected field [remote_cluster]" for CCS (RCS 1.0) when using API key that references `remote_cluster` {es-pull}112226[#112226]
* Fix connection timeout for `OpenIdConnectAuthenticator` get Userinfo {es-pull}112230[#112230]
Vector Search::
* Fix `NullPointerException` when doing knn search on empty index without dims {es-pull}111756[#111756] (issue: {es-issue}111733[#111733])
* Speed up dense/sparse vector stats {es-pull}111729[#111729] (issue: {es-issue}111715[#111715])