A failed NRT reader open (DirectoryReader.open(IndexWriter), openIfChanged over the writer, or a SearcherManager/ReaderManager refresh) whose IndexWriterConfig leaf sorter throws never releases the IndexFileDeleter reference it took, so the still-open writer keeps every segment file of that point in time on disk for the rest of its life

Finding LFV-1067Area: core / NRT reader open from IndexWriter (StandardDirectoryReader.open(IndexWriter, ...) failure path, with an IndexWriterConfig leaf sorter, LUCENE-9507)Priority: lowImpact: minorNovelty: new

Finding page for LFV-1067

AI-assisted finding, confirmed by automated checks, not yet triaged by a human Lucene committer. This site files nothing upstream.

What it shows

You set a leaf sorter on IndexWriterConfig and open an NRT reader from the writer; if the sorter throws while the new reader sorts its leaves, you expect only that exception and a writer left as it was, because the open's own catch releases the segment readers it took and IndexWriter.getReader keeps the writer usable after a non-Error exception. Instead the open has already called writer.incRefDeleter(segmentInfos), which only the returned reader's close undoes, and no reader is returned, so the writer keeps every segment file of that point in time referenced for the rest of its life: after forceMerge(1) and commit, the files of the merged-away segments _0 and _1 are still in the directory.

Expected vs actual

Expected: After the failed open, forceMerge(1) and commit(), no file of _0 or _1 is left; only _2's files and segments_3 remain.
Actual: AssertionError: file of a merged-away segment was not deleted: _0.cfe; all 6 files of _0 and _1 remain, assertions on and off, on main and 10.6.0.

Files

src/org/apache/lucene/index/TestNrtOpenLeafSorterFailureReleasesFiles.java (JUnit test)

Run it

Tested on Lucene main e1faeba08bde (2026-11-02): still reproduces.
Tested on Lucene main 645d4fe14a20 (2026-10-28): reproduced.
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 off (production-like), seed DEADBEEF
java -da -dsa -Dtests.asserts=false --add-modules jdk.incubator.vector --enable-native-access=ALL-UNNAMED -Dtests.seed=DEADBEEF \
     -cp "classes:$CP" org.junit.platform.console.ConsoleLauncher \
     execute --select-class org.apache.lucene.index.TestNrtOpenLeafSorterFailureReleasesFiles

Gradle 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).

Files shown on the finding page

TestNrtOpenLeafSorterFailureReleasesFiles.java

Formal models and scripts this finding's part of the hunt used: reader-open-leaf-sorter-failure-leaks

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 2026-11-04. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.