Skip to content

[Subscription] Flush lingering batch before WAL gap retry #18785

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jt2594838 merged 1 commit into from
Sep 30, 2026
Merged

[Subscription] Flush lingering batch before WAL gap retry #18785

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions ...ain/java/org/apache/iotdb/db/subscription/broker/consensus/ConsensusPrefetchingQueue.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1507,6 +1507,10 @@
}
if (batchResult != MaterializationResult.SUCCESS) {
if (batchResult == MaterializationResult.WAL_GAP) {
if (!lingerBatch.isEmpty() && !flushBatch(lingerBatch, observedSeekGeneration)) {
resetRoundStateForSeek(seekGeneration.get());
return PrefetchRoundResult.rescheduleNow();
}
return PrefetchRoundResult.rescheduleAfter(WAL_GAP_RETRY_SLEEP_MS);
}
if (batchResult == MaterializationResult.MEMORY_BLOCKED) {
Expand Down Expand Up @@ -2100,7 +2104,7 @@
walReplayFailureMessage =
String.format(
DataNodePipeMessages
.MESSAGE_CONSENSUSPREFETCHINGQUEUE_WAL_REPLAY_CANNOT_RECOVER_SEARCH_INDEXES_70781B22,

Check warning on line 2107 in iotdb-core/datanode/src/main/java/org/apache/iotdb/db/subscription/broker/consensus/ConsensusPrefetchingQueue.java

View check run for this annotation

SonarQubeCloud / SonarCloud Code Analysis

Line is longer than 100 characters (found 103). 


See more on https://sonarcloud.io/project/issues?id=apache_iotdb&issues=AaDyjP5mBxenDX0ixV_x&open=AaDyjP5mBxenDX0ixV_x&pullRequest=18785
this,
expectedSearchIndex,
actualSearchIndex,
Expand Down
88 changes: 88 additions & 0 deletions ...java/org/apache/iotdb/db/subscription/broker/consensus/ConsensusPrefetchingQueueTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1609,6 +1609,94 @@ public void testWalReplayCountsOnlyUnavailableSearchIndexes() throws Exception {
}
}

@Test
public void testPendingWalGapFlushesLingerBatchBeforeRetry() throws Exception {
final String originalSystemDir = IoTDBDescriptor.getInstance().getConfig().getSystemDir();
final int originalBatchMaxWalEntries =
CommonDescriptor.getInstance().getConfig().getSubscriptionConsensusBatchMaxWalEntries();
final int originalBatchMaxDelay =
CommonDescriptor.getInstance().getConfig().getSubscriptionConsensusBatchMaxDelayInMs();
final File systemDir = temporaryFolder.newFolder("pending-gap-flush-linger-batch");
ConsensusPrefetchingQueue queue = null;
try {
CommonDescriptor.getInstance().getConfig().setSubscriptionConsensusBatchMaxWalEntries(2);
CommonDescriptor.getInstance().getConfig().setSubscriptionConsensusBatchMaxDelayInMs(60_000);

final DataRegionId regionId = new DataRegionId(13);
final WALNode walNode = mock(WALNode.class);
when(walNode.getCurrentSearchIndex()).thenReturn(5L);
when(walNode.getLogDirectory()).thenReturn(systemDir);
final IoTConsensusServerImpl serverImpl = mock(IoTConsensusServerImpl.class);
when(serverImpl.getConsensusReqReader()).thenReturn(walNode);
when(serverImpl.getWriterSafeFrontierTracker()).thenReturn(new WriterSafeFrontierTracker());

final ConsensusLogToTabletConverter converter = mock(ConsensusLogToTabletConverter.class);
when(converter.convert(any())).thenReturn(Collections.singletonList(createTablet()));
when(converter.getDatabaseName()).thenReturn("db");

queue =
new ConsensusPrefetchingQueue(
"consumerGroup",
"topic",
TopicConstant.ORDER_MODE_LEADER_ONLY_VALUE,
regionId,
serverImpl,
new SubscriptionWalRetentionPolicy(
"topic",
SubscriptionWalRetentionPolicy.UNBOUNDED,
SubscriptionWalRetentionPolicy.UNBOUNDED),
converter,
newCommitManager(systemDir),
new RegionProgress(Collections.emptyMap()),
1L,
1L,
true) {
@Override
protected ProgressWALIterator createSubscriptionWALIterator(
final long startSearchIndex) {
final Iterator<IndexedConsensusRequest> walEntries =
Arrays.asList(
createRequest(1L),
createRequest(2L),
createRequest(3L),
createRequest(4L),
createRequest(5L))
.iterator();
final ProgressWALIterator iterator = mock(ProgressWALIterator.class);
when(iterator.hasNext()).thenAnswer(ignored -> walEntries.hasNext());
when(iterator.next()).thenAnswer(ignored -> walEntries.next());
return iterator;
}
};
queue.setSubscriptionMemoryManager(new SubscriptionMemoryManager(16L * 1024 * 1024));

assertNull(queue.poll("consumer"));
assertTrue(pendingEntries(queue).offer(createRequest(5L)));

queue.drivePrefetchOnce();

assertEquals(3L, queue.getCurrentReadSearchIndex());
assertEquals(1, queue.getPrefetchedEventCount());
final SubscriptionEvent event = queue.poll("consumer");
assertNotNull(event);
assertEquals(
SubscriptionPollResponseType.TABLETS.getType(),
event.getCurrentResponse().getResponseType());
assertTrue(queue.ack("consumer", event.getCommitContext()));
} finally {
if (queue != null) {
queue.close();
}
CommonDescriptor.getInstance()
.getConfig()
.setSubscriptionConsensusBatchMaxWalEntries(originalBatchMaxWalEntries);
CommonDescriptor.getInstance()
.getConfig()
.setSubscriptionConsensusBatchMaxDelayInMs(originalBatchMaxDelay);
IoTDBDescriptor.getInstance().getConfig().setSystemDir(originalSystemDir);
}
}

@Test
public void testWalReplayFailsCriticallyWhenGapRemainsUnavailable() throws Exception {
final String originalSystemDir = IoTDBDescriptor.getInstance().getConfig().getSystemDir();
Expand Down
Loading