An NRT replica's commit() can store in VERSION_KEY the version of a newer NRT point installed during its fsync, so the commit claims a version ahead of the SegmentInfos it holds

Finding LFV-260Area: replicator / NRT segment replication (ReplicaNode.commit commit user data, PrimaryNode.getLastCommitVersion)Priority: lowImpact: silent wrong resultsNovelty: new

Finding page for LFV-260

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 Lucene's NRT segment replication and call commit() on a replica while the primary keeps publishing NRT points, and during the commit's fsync the replica installs a newer point (version 5, documents [d1, d2]). You expect the commit's VERSION_KEY to name the version whose content it holds (version 3, [d1]), or at worst an older one, because the module uses that value as the place to resume replaying operations and PrimaryNode.commit accepts only a lagging value ('replaying more ops than necessary'). Instead commit() writes the SegmentInfos it captured before the fsync (version 3) but reads VERSION_KEY from the searcher that is current after it (5), so a primary started on that commit reports getLastCommitVersion()=5 while it holds only d1, and a replay from version 5 skips d2.

The test shows that ReplicaNode.commit() stores in VERSION_KEY the version of a newer NRT point installed while the commit fsyncs, not the version of the SegmentInfos it commits.

Expected vs actual

Expected VERSION_KEY = the committed SegmentInfos version 3; actual: 5 (expected:<3> but was:<5>), with -ea and -da.

Files

• src/org/apache/lucene/replicator/nrt/TestReplicaCommitVersionKey.java (JUnit test)

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=$CP:$LUCENE/lucene/facet/build/libs/lucene-facet-$V.jar
CP=$CP:$LUCENE/lucene/replicator/build/libs/lucene-replicator-$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 DEADBEEF
java -ea -esa --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.replicator.nrt.TestReplicaCommitVersionKey

The module(s) facet, replicator also need their third-party runtime jars; ./gradlew -p lucene/<module> dependencies lists them.

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).

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.