kNN vector queries size each segment's result heap from the requested k, not from what the segment can return, so k = Integer.MAX_VALUE throws IllegalArgumentException and k = 2^28 runs out of a 2 GiB heap on a 3-document index (core KnnFloat/KnnByteVectorQuery without a filter: regression in 9.10.0; join DiversifyingChildren*KnnVectorQuery: since 9.8.0)
Finding LFV-454Area: search / kNN vector queries (per-segment collector sizing: AbstractKnnVectorQuery, TopKnnCollectorManager, TopKnnCollector; join DiversifyingNearestChildrenKnnCollectorManager)Priority: lowImpact: crashNovelty: new
AI-assisted finding, confirmed by automated checks, not yet triaged by a human Lucene committer. This site files nothing upstream.
What it shows
You run a KnnFloatVectorQuery with a k larger than the index, for example k = Integer.MAX_VALUE to mean 'all vectors', and expect every vector back, as k = 1000 over 3 vectors gives, and as 9.9.2 gives for any k. Instead each segment's collector is built with a heap sized from k (k + 1 longs on a single-segment index) before the segment's vector count is looked at: k = Integer.MAX_VALUE throws IllegalArgumentException('initialCapacity must be > 0 and < 2147483630; got: 2147483647') from TernaryLongHeap, and k = 2^28 over a 3-document index throws OutOfMemoryError with a 2 GiB heap. Lucene used to cap k at the segment's vector count ('bound k by total number of vectors to prevent oversizing data structures'); a search speed-up change dropped that cap in 9.10.0, and the join DiversifyingChildren*KnnVectorQuery never had it.
The test shows that KnnFloatVectorQuery without a filter, k = Integer.MAX_VALUE, over a 3-vector single-segment index sizes the per-segment collector heap from k (TestKnnQueryKMaxValue#testKMaxValue).
Expected vs actual
Expected 3 hits (k is "the number of documents to find"); actual IllegalArgumentException 'initialCapacity must be > 0 and < 2147483630; got: 2147483647' from TernaryLongHeap, with -ea and -da.
Files
Run it
- Tested on Lucene main e1faeba08bde (2026-11-02): still reproduces.
- Tested on Lucene main 645d4fe14a20 (2026-10-28): reproduced.
- Recorded outcome of run 1 on Lucene main 645d4fe14a20: 1 test(s) run, 1 failed.
- How to read the outcome: a failing test is the finding reproducing (the failure message states expected against actual).
- Seeds: a rerun with the same seed replays the same random choices. Assertions off (-da -dsa -Dtests.asserts=false) is the production-like setting.
# Lucene checkout at the tested commit, built with: ./gradlew --no-daemon assemble (JDK 25)
LUCENE=/path/to/lucene; V=12.0.0-SNAPSHOT
CP=$LUCENE/lucene/core/build/libs/lucene-core-$V.jar
CP=$CP:$LUCENE/lucene/codecs/build/libs/lucene-codecs-$V.jar
CP=$CP:$LUCENE/lucene/test-framework/build/libs/lucene-test-framework-$V.jar
CP=$CP:randomizedtesting-runner-2.9.2.jar:junit-4.13.2.jar:hamcrest-3.0.jar # test-framework dependencies (Maven Central)
CP=$CP:junit-platform-console-standalone-6.1.3.jar # JUnit Platform console launcher (Maven Central)
javac -d classes -cp "$CP" $(find src -name '*.java')
# run 1: assertions on, seed 5EED0454
java -ea -esa --add-modules jdk.incubator.vector --enable-native-access=ALL-UNNAMED -Dtests.seed=5EED0454 \
-cp "classes:$CP" org.junit.platform.console.ConsoleLauncher \
execute --select-class org.apache.lucene.search.TestKnnQueryKMaxValueGradle alternative: copy the test into the matching module's src/test/ tree of the checkout and run ./gradlew -p lucene/<module> test --tests <class> -Ptests.seed=<seed> (add -Ptests.asserts=false for a run without assertions).
Feedback
Open an issue in lucene-ai-bug-hunt. Please do not open issues or pull requests in apache/lucene about this finding unless a Lucene committer agrees it should be filed.
Snapshot of 2026-11-03. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.