Fix ES|QL search context creation to use correct results type#137994
Merged
jimczi merged 2 commits intoelastic:mainfrom Nov 13, 2025
Merged
Fix ES|QL search context creation to use correct results type#137994jimczi merged 2 commits intoelastic:mainfrom
jimczi merged 2 commits intoelastic:mainfrom
Conversation
The ES|QL search context was being created with an incorrect results type, which unintentionally disabled the inner executor in the Lucene index searcher. This hadn’t caused issues previously because parallelisation happens outside the index searcher for most queries, but KNN queries rely on the internal executor for proper parallel rewrite execution. This change ensures that the correct results type is used so KNN queries executed via ES|QL can leverage the inner executor and parallelise their rewrite phase correctly.
jimczi
added
>non-issue
:Search Relevance/ES|QL
elasticsearchmachine
added
the
Team:Search Relevance
Collaborator
|
Pinging @elastic/es-search-relevance (Team:Search Relevance) |
carlosdelest
approved these changes
Nov 13, 2025
Member
carlosdelest
left a comment
carlosdelest
left a comment
There was a problem hiding this comment.
LGTM, thanks for fixing this!
szybia
added a commit
to szybia/elasticsearch
that referenced
this pull request
Nov 13, 2025
…-json * upstream/main: (158 commits) Cleanup files from repo root folder (elastic#138030) Implement OpenShift AI integration for chat completion, embeddings, and reranking (elastic#136624) Optimize AsyncSearchErrorTraceIT to avoid failures (elastic#137716) Removes support for null TransportService in RemoteClusterService (elastic#137939) Mute org.elasticsearch.index.mapper.DateFieldMapperTests testSortShortcuts elastic#138018 rest-api-spec: fix type of enums (elastic#137521) Update Gradle wrapper to 9.2.0 (elastic#136155) Add RCS Strong Verification Documentation (elastic#137822) Use docvalue skippers on dimension fields (elastic#137029) Introduce INDEX_SHARD_COUNT_FORMAT (elastic#137210) Mute org.elasticsearch.xpack.inference.integration.AuthorizationTaskExecutorIT testCreatesChatCompletion_AndThenCreatesTextEmbedding elastic#138012 Fix ES|QL search context creation to use correct results type (elastic#137994) Improve Snapshot Logging (elastic#137470) Support extra output field in TOP function (elastic#135434) Remove NumericDoubleValues class (elastic#137884) [ML] Fix ML calendar event update scalability issues (elastic#136886) Task may be unregistered outside of the trace context in exceptional cases. (elastic#137865) Refine workaround for S3 repo analysis known issue (elastic#138000) Additional DEBUG logging on authc failures (elastic#137941) Cleanup index resolution (elastic#137867) ...
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The ES|QL search context was being created with an incorrect results type, which unintentionally disabled the inner executor in the Lucene index searcher. This hadn’t caused issues previously because parallelisation happens outside the index searcher for most queries, but KNN queries rely on the internal executor for proper parallel rewrite execution.
This change ensures that the correct results type is used so KNN queries executed via ES|QL can leverage the inner executor and parallelise their rewrite phase correctly.