An NRT replica acts on a late notification from an older primary generation: ReplicaNode.maybeNewPrimary deletes the current primary's pre-copied merged segment before its only check, an assert, so the replica's cut-over to the merged point fails or the segment is copied again (with or without assertions), and with assertions off lastPrimaryGen also moves back

Finding LFV-270Area: replicator / NRT segment replication (ReplicaNode.newNRTPoint, maybeNewPrimary, launchPreCopyMerge: primaryGen fencing)Priority: lowImpact: crashNovelty: new

Finding page for LFV-270

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 fail over cleanly: the old primary (generation 1) is closed, then replica A is promoted to primary (generation 2), and replica B follows A and pre-copies A's merged segment. A last notification from the old primary, delayed in transit, then reaches B; you expect B to ignore it, as it already ignores out-of-order older NRT versions, because the primary generation only ever grows. Instead ReplicaNode.maybeNewPrimary treats the older generation as a primary change and deletes A's pre-copied merged files before its only check (an assert), so B's cut-over to A's merged point fails with '_3.si cannot be read - it may be missing or corrupt' or B copies the merged segment again, and with assertions off B also sets lastPrimaryGen back to 1.

The test shows that an NRT replica following primaryGen 2 deletes that primary's pre-copied merged segment when a late newNRTPoint(1, olderVersion) from the older generation arrives (ReplicaNode.maybeNewPrimary deletes pendingMergeFiles before its only check, an assert).

Expected vs actual

Expected: the stale call is ignored or rejected and [_2.cfe, _2.cfs, _2.si] stay in the replica's directory. Actual (-ea and -da): the three files are gone.

Files

Run it

# 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.TestReplicaLateOldPrimaryNotification

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 2026-11-03. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.