Disable max score optimization for queries with unbounded max scores #41361
Conversation
This file contains 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
Lucene 8 has the ability to skip blocks of non-competitive documents. However some queries don't track their maximum score (`script_score`, `span`, ...) so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down some boolean queries if other clauses have bounded max scores. This commit disables the max score optimization when we detect a mandatory scoring clause with unbounded max scores. Optional clauses are not checked since they can still skip documents when the unbounded clause is after the current document.
jimczi
added
>enhancement
:Search/Search
|
Pinging @elastic/es-search |
romseygeek
suggested changes
Apr 22, 2019
| hasInfMaxScore = true; | ||
| } else if (query instanceof ESToParentBlockJoinQuery) { | ||
| ESToParentBlockJoinQuery q = (ESToParentBlockJoinQuery) query; | ||
| hasInfMaxScore = q.getScoreMode() != org.apache.lucene.search.join.ScoreMode.None; |
There was a problem hiding this comment.
This should be |= I think, otherwise if you have a conjunction with both SpanQuery and ESToParentBlockJoinQuery, you might end up un-setting the boolean here.
There was a problem hiding this comment.
good catch, I pushed 0f4de62
|
run elasticsearch-ci/1 |
|
This cannot be backported to 7x yet since this pr relies on code that is only in Lucene 8x (QueryVisitor) and 7x is still on Lucene 8.0 |
jasontedor
added a commit
to jasontedor/elasticsearch
that referenced
this pull request
Apr 26, 2019
…s-in-all-the-places * elastic/master: (70 commits) Remove experimental label froms script_score query (elastic#41572) [Ml-Dataframe] Update URLs in Data frame client java doc (elastic#41539) Reenable bwc Tests in master (elastic#41540) Fix search_as_you_type's sub-fields to pick their names from the full path of the root field (elastic#41541) Remove search analyzers from DocumentFieldMappers (elastic#41484) Update community client and integration docs (elastic#41513) Remove Version.V_6_x_x constants use in security (elastic#41185) Mute testDriverConfigurationWithSSLInURL Remove dedicated SSL network write buffer (elastic#41283) Disable max score optimization for queries with unbounded max scores (elastic#41361) [DOCS] Explicitly set section IDs for Asciidoctor migration (elastic#41547) [ML] add multi node integ tests for data frames (elastic#41508) [Docs] Fix common word repetitions (elastic#39703) [DOCS] Note TESTRESPONSE can't be used immediately after TESTSETUP (elastic#41542) Update configuring-ldap-realm.asciidoc (elastic#40427) Fixed very small typo in date (elastic#41398) Refactor GeoHashUtils (elastic#40869) Make 0 as invalid value for `min_children` in `has_child` query (elastic#41347) field_caps: adapt bwc version after backport (elastic#41427) Remove Exists Check from S3 Repository Deletes (elastic#40931) ...
akhil10x5
pushed a commit
to akhil10x5/elasticsearch
that referenced
this pull request
May 2, 2019
…lastic#41361) Lucene 8 has the ability to skip blocks of non-competitive documents. However some queries don't track their maximum score (`script_score`, `span`, ...) so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down some boolean queries if other clauses have bounded max scores. This commit disables the max score optimization when we detect a mandatory scoring clause with unbounded max scores. Optional clauses are not checked since they can still skip documents when the unbounded clause is after the current document.
jimczi
added a commit
that referenced
this pull request
May 23, 2019
…41361) Lucene 8 has the ability to skip blocks of non-competitive documents. However some queries don't track their maximum score (`script_score`, `span`, ...) so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down some boolean queries if other clauses have bounded max scores. This commit disables the max score optimization when we detect a mandatory scoring clause with unbounded max scores. Optional clauses are not checked since they can still skip documents when the unbounded clause is after the current document.
gurkankaymak
pushed a commit
to gurkankaymak/elasticsearch
that referenced
this pull request
May 27, 2019
…lastic#41361) Lucene 8 has the ability to skip blocks of non-competitive documents. However some queries don't track their maximum score (`script_score`, `span`, ...) so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down some boolean queries if other clauses have bounded max scores. This commit disables the max score optimization when we detect a mandatory scoring clause with unbounded max scores. Optional clauses are not checked since they can still skip documents when the unbounded clause is after the current document.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Lucene 8 has the ability to skip blocks of non-competitive documents.
However some queries don't track their maximum score (
script_score,span, ...)so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down
some boolean queries if other clauses have bounded max scores. This commit disables
the max score optimization when we detect a mandatory scoring clause with unbounded
max scores. Optional clauses are not checked since they can still skip documents
when the unbounded clause is after the current document.