Skip to content

KAFKA-20330: Ack handling improvement on broker restart #21824

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
AndrewJSchofield merged 3 commits into from
Mar 30, 2026
Merged

KAFKA-20330: Ack handling improvement on broker restart #21824

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
8b53e4b
KAFKA-20330: Ack handling improvement on broker restart
AndrewJSchofield Mar 19, 2026
1f704d7
Remove debug statements
AndrewJSchofield Mar 20, 2026
2c8cfb3
Adjust ack processing on close and tests
AndrewJSchofield Mar 20, 2026
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
141 changes: 141 additions & 0 deletions ...-integration-tests/src/test/java/org/apache/kafka/clients/consumer/ShareConsumerTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,8 +48,10 @@
import org.apache.kafka.common.errors.InvalidConfigurationException;
import org.apache.kafka.common.errors.InvalidRecordStateException;
import org.apache.kafka.common.errors.InvalidTopicException;
import org.apache.kafka.common.errors.NotLeaderOrFollowerException;
import org.apache.kafka.common.errors.RecordDeserializationException;
import org.apache.kafka.common.errors.SerializationException;
import org.apache.kafka.common.errors.ShareSessionNotFoundException;
import org.apache.kafka.common.errors.UnknownTopicIdException;
import org.apache.kafka.common.errors.WakeupException;
import org.apache.kafka.common.header.Header;
Expand Down Expand Up @@ -1301,6 +1303,145 @@ public void testConsumerCloseOnBrokerShutdown() {
"Consumer close should not wait for full timeout when broker is already shut down");
}

@ClusterTest
public void testLeaderRestartWithoutLeadershipChangeExplicitAcknowledgementSync() {
alterShareAutoOffsetReset("group1", "earliest");
try (Producer<byte[], byte[]> producer = createProducer();
ShareConsumer<byte[], byte[]> shareConsumer = createShareConsumer("group1",
Map.of(ConsumerConfig.SHARE_ACKNOWLEDGEMENT_MODE_CONFIG, EXPLICIT))) {

AtomicBoolean callbackCalled = new AtomicBoolean(false);
shareConsumer.setAcknowledgementCommitCallback((offsetsByTopicPartition, exception) -> {
assertInstanceOf(NotLeaderOrFollowerException.class, exception);
callbackCalled.set(true);
});

ProducerRecord<byte[], byte[]> record = new ProducerRecord<>(tp.topic(), tp.partition(), null, "key".getBytes(), "value".getBytes());
shareConsumer.subscribe(Set.of(tp.topic()));

producer.send(record);
producer.flush();

ConsumerRecords<byte[], byte[]> records = shareConsumer.poll(Duration.ofMillis(20000));
assertEquals(1, records.count());
ConsumerRecord<byte[], byte[]> consumerRecord = records.iterator().next();

// Shutdown the broker
assertEquals(1, cluster.brokers().size());
KafkaBroker broker = cluster.brokers().get(0);
cluster.shutdownBroker(0);

broker.awaitShutdown();

// Restart the broker
cluster.startBroker(0);

shareConsumer.acknowledge(consumerRecord);
Map<TopicIdPartition, Optional<KafkaException>> commitResult = shareConsumer.commitSync(Duration.ofMillis(30000));
assertEquals(1, commitResult.size());
TopicIdPartition tidp = commitResult.keySet().iterator().next();
assertTrue(commitResult.get(tidp).isPresent());
assertInstanceOf(NotLeaderOrFollowerException.class, commitResult.get(tidp).get());

assertTrue(callbackCalled.get());
}
}

@ClusterTest
public void testLeaderRestartWithoutLeadershipChangeExplicitAcknowledgementAsync() {
alterShareAutoOffsetReset("group1", "earliest");
try (Producer<byte[], byte[]> producer = createProducer();
ShareConsumer<byte[], byte[]> shareConsumer = createShareConsumer("group1",
Map.of(ConsumerConfig.SHARE_ACKNOWLEDGEMENT_MODE_CONFIG, EXPLICIT))) {

AtomicBoolean callbackCalled = new AtomicBoolean(false);
shareConsumer.setAcknowledgementCommitCallback((offsetsByTopicPartition, exception) -> {
assertInstanceOf(NotLeaderOrFollowerException.class, exception);
callbackCalled.set(true);
});

ProducerRecord<byte[], byte[]> record = new ProducerRecord<>(tp.topic(), tp.partition(), null, "key".getBytes(), "value".getBytes());
shareConsumer.subscribe(Set.of(tp.topic()));

producer.send(record);
producer.flush();

ConsumerRecords<byte[], byte[]> records = shareConsumer.poll(Duration.ofMillis(20000));
assertEquals(1, records.count());
ConsumerRecord<byte[], byte[]> consumerRecord = records.iterator().next();

// Shutdown the broker
assertEquals(1, cluster.brokers().size());
KafkaBroker broker = cluster.brokers().get(0);
cluster.shutdownBroker(0);

broker.awaitShutdown();

// Restart the broker
cluster.startBroker(0);

shareConsumer.acknowledge(consumerRecord);
shareConsumer.commitAsync();

int maxRetries = 15;
int retries = 0;
while (retries < maxRetries) {
shareConsumer.poll(Duration.ofMillis(2000));
if (callbackCalled.get()) {
break;
}
retries++;
}

assertTrue(callbackCalled.get());
}
}

@ClusterTest
public void testLeaderRestartWithoutLeadershipChangeImplicitAcknowledgement() {
alterShareAutoOffsetReset("group1", "earliest");
try (Producer<byte[], byte[]> producer = createProducer();
ShareConsumer<byte[], byte[]> shareConsumer = createShareConsumer("group1")) {

AtomicBoolean callbackCalled = new AtomicBoolean(false);
shareConsumer.setAcknowledgementCommitCallback((offsetsByTopicPartition, exception) -> {
assertInstanceOf(ShareSessionNotFoundException.class, exception);
callbackCalled.set(true);
});

ProducerRecord<byte[], byte[]> record = new ProducerRecord<>(tp.topic(), tp.partition(), null, "key".getBytes(), "value".getBytes());
shareConsumer.subscribe(Set.of(tp.topic()));

producer.send(record);
producer.flush();

ConsumerRecords<byte[], byte[]> records = shareConsumer.poll(Duration.ofMillis(20000));
assertEquals(1, records.count());

// Shutdown the broker
assertEquals(1, cluster.brokers().size());
KafkaBroker broker = cluster.brokers().get(0);
cluster.shutdownBroker(0);

broker.awaitShutdown();

// Restart the broker
cluster.startBroker(0);

int maxRetries = 15;
int retries = 0;
while (retries < maxRetries) {
shareConsumer.poll(Duration.ofMillis(2000));
if (callbackCalled.get()) {
break;
}
retries++;
}

assertTrue(callbackCalled.get());
}
}

@ClusterTests({
@ClusterTest(serverProperties = {
@ClusterConfigProperty(key = GroupCoordinatorConfig.CONSUMER_GROUP_ASSIGNMENT_INTERVAL_MS_CONFIG, value = "0")
Expand Down
Loading
Loading