Add query planner Phase 2: logical planning and cost estimation #19244
Conversation
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
Introduces LogicalPlanBuilder to convert QueryBuilder to QueryPlanNode tree with multi-dimensional cost tracking (lucene_docs, cpu, memory, io). Adds CostEstimator leveraging Lucene's scorer costs across index leaves with OpenSearch-specific heuristics. Implements volatile cost caching with double-checked locking for thread safety. Supports Boolean, Term, Range, and Match queries with QueryBuilder context preservation. Includes feature flag (opensearch.experimental.feature.query_planner.enabled) for gradual rollout and QueryPlannerIntegration helper for search pipeline integration. Bounded must_not penalty prevents cost explosion on large document sets. Signed-off-by: Atri Sharma <[email protected]>
Signed-off-by: Atri Sharma <[email protected]>
Signed-off-by: Atri Sharma <[email protected]>
atris
requested review from
anasalkouz,
andrross,
ashking94,
Bukhtawar,
CEHENKLE,
cwperks,
dbwiddis,
gbbafna,
jed326,
kotwanikunal,
mch2,
msfroh,
owaiskazi19,
reta,
Rishikesh1159,
sachinpkale,
saratvemulapalli,
shwetathareja,
sohami and
a team
as code owners
|
@andrross @rishabhmaurya Please review |
|
❌ Gradle check result for 75e057f: FAILURE Please examine the workflow log, locate, and copy-paste the failure(s) below, then iterate to green. Is the failure a flaky test unrelated to your change? |
|
❌ Gradle check result for 75e057f: FAILURE Please examine the workflow log, locate, and copy-paste the failure(s) below, then iterate to green. Is the failure a flaky test unrelated to your change? |
This was referenced Sep 8, 2025
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.
Following up on Phase 1 (PR #19060), this adds the cost estimation
infrastructure needed for query optimization (RFC #18906).
What this does:
Builds a tree of QueryPlanNode objects from QueryBuilder
Estimates query costs by summing Lucene scorer costs across all segments
Tracks CPU/memory/IO costs alongside document counts
Preserves field names and query metadata (lost when converting to Lucene)
The main insight is using Lucene's existing scorer.cost() aggregated
properly across index leaves, then adding our own heuristics for
things Lucene doesn't track.
Notable implementation details:
BooleanPlanNode caps must_not penalties so costs don't explode
Percentage minimumShouldMatch now works (e.g. "50%")
Feature flagged: -Dopensearch.experimental.feature.query_planner.enabled
No user-visible changes yet - this is just the foundation for Phase 3
where we'll actually optimize queries based on these costs.
Signed-off-by: Atri Sharma [email protected]
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.