Skip to content

Do not clear offsets after failure to commit offsets due to rebalance exception #6346

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
graytaylor0 merged 1 commit into from
Dec 15, 2025
+91 −1
Merged

Do not clear offsets after failure to commit offsets due to rebalance exception #6346

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
7 changes: 6 additions & 1 deletion .../src/main/java/org/opensearch/dataprepper/plugins/kafka/consumer/KafkaCustomConsumer.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.clients.consumer.OffsetAndMetadata;
import org.apache.kafka.common.errors.RebalanceInProgressException;
import org.apache.kafka.common.header.Header;
import org.apache.kafka.common.header.Headers;
import org.apache.kafka.common.TopicPartition;
Expand Down Expand Up @@ -354,11 +355,15 @@ private void commitOffsets(boolean forceCommit) {
offsetsToCommit.forEach(((partition, offset) -> updateCommitCountMetric(partition, offset)));
try {
consumer.commitSync(offsetsToCommit);
lastCommitTime = currentTimeMillis;
} catch (final RebalanceInProgressException ex) {
LOG.error("Failed to commit offsets in topic {} due to rebalance in progress", topicName, ex);
return;
} catch (Exception e) {

@dlvenable dlvenable Dec 9, 2025

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Currently, this exception set the lastCommitTime. But, now we don't. Is that ok?

@graytaylor0 graytaylor0 Dec 9, 2025

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It seems like it was incorrectly saying we committed offsets when we didn't if we get an exception

@kkondaka kkondaka Dec 9, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dlvenable it is correct because the last commit time should be set only if commit really happens.

LOG.error("Failed to commit offsets in topic {}", topicName, e);
}

offsetsToCommit.clear();
lastCommitTime = currentTimeMillis;
}
}

Expand Down
85 changes: 85 additions & 0 deletions .../test/java/org/opensearch/dataprepper/plugins/kafka/consumer/KafkaCustomConsumerTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,7 @@
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.clients.consumer.OffsetAndMetadata;
import org.apache.kafka.common.TopicPartition;
import org.apache.kafka.common.errors.RebalanceInProgressException;
import org.apache.kafka.common.errors.RecordDeserializationException;
import org.junit.jupiter.api.Assertions;
import org.junit.jupiter.api.BeforeEach;
Expand Down Expand Up @@ -60,6 +61,8 @@
import static org.junit.jupiter.api.Assertions.assertTrue;
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.ArgumentMatchers.anyInt;
import static org.mockito.ArgumentMatchers.anyMap;
import static org.mockito.Mockito.doThrow;
import static org.mockito.Mockito.doAnswer;
import static org.mockito.Mockito.mock;
import static org.mockito.Mockito.when;
Expand Down Expand Up @@ -589,6 +592,88 @@ public void testAwsGlueErrorWithAcknowledgements() throws Exception {
});
}

@Test
public void testCommitOffsets_RebalanceInProgressException_DoesNotClearOffsets() throws Exception {
String topic = topicConfig.getName();
TopicPartition topicPartition = new TopicPartition(topic, testPartition);

when(topicConfig.getSerdeFormat()).thenReturn(MessageFormat.PLAINTEXT);
when(topicConfig.getAutoCommit()).thenReturn(false);
when(topicConfig.getCommitInterval()).thenReturn(Duration.ofMillis(0));

consumer = createObjectUnderTest("plaintext", false);
consumer.onPartitionsAssigned(List.of(topicPartition));

consumerRecords = createPlainTextRecords(topic, 100L);
when(kafkaConsumer.poll(any(Duration.class))).thenReturn(consumerRecords);

doThrow(new RebalanceInProgressException("Rebalance in progress"))
.when(kafkaConsumer).commitSync(anyMap());

consumer.consumeRecords();

Map<TopicPartition, OffsetAndMetadata> offsetsBeforeCommit = new HashMap<>(consumer.getOffsetsToCommit());
Assertions.assertFalse(offsetsBeforeCommit.isEmpty(), "Offsets should be populated after consuming records");
Assertions.assertEquals(102L, offsetsBeforeCommit.get(topicPartition).offset());

Thread testThread = new Thread(() -> {
try {
java.lang.reflect.Method method = consumer.getClass().getDeclaredMethod("commitOffsets", boolean.class);
method.setAccessible(true);
method.invoke(consumer, true);
} catch (Exception e) {
throw new RuntimeException(e);
}
});
testThread.start();
testThread.join(5000);

Map<TopicPartition, OffsetAndMetadata> offsetsAfterFailedCommit = consumer.getOffsetsToCommit();
Assertions.assertFalse(offsetsAfterFailedCommit.isEmpty(),
"Offsets should NOT be cleared after RebalanceInProgressException");
Assertions.assertEquals(offsetsBeforeCommit.get(topicPartition).offset(),
offsetsAfterFailedCommit.get(topicPartition).offset(),
"Offset value should remain unchanged for retry after rebalance completes");
}

@Test
public void testCommitOffsets_OtherException_ClearsOffsets() throws Exception {
String topic = topicConfig.getName();
TopicPartition topicPartition = new TopicPartition(topic, testPartition);

when(topicConfig.getAutoCommit()).thenReturn(false);
when(topicConfig.getCommitInterval()).thenReturn(Duration.ofMillis(0));

consumer = createObjectUnderTest("plaintext", false);
consumer.onPartitionsAssigned(List.of(topicPartition));

consumerRecords = createPlainTextRecords(topic, 100L);
when(kafkaConsumer.poll(any(Duration.class))).thenReturn(consumerRecords);

doThrow(new RuntimeException("Generic commit failure"))
.when(kafkaConsumer).commitSync(anyMap());

consumer.consumeRecords();

Assertions.assertFalse(consumer.getOffsetsToCommit().isEmpty(),
"Offsets should be populated after consuming records");

Thread testThread = new Thread(() -> {
try {
java.lang.reflect.Method method = consumer.getClass().getDeclaredMethod("commitOffsets", boolean.class);
method.setAccessible(true);
method.invoke(consumer, true);
} catch (Exception e) {
}
});
testThread.start();
testThread.join(5000);

Map<TopicPartition, OffsetAndMetadata> offsetsAfterFailedCommit = consumer.getOffsetsToCommit();
Assertions.assertTrue(offsetsAfterFailedCommit.isEmpty(),
"Offsets should be cleared after non-rebalance exception");
}

private ConsumerRecords createPlainTextRecords(String topic, final long startOffset) {
Map<TopicPartition, List<ConsumerRecord>> records = new HashMap<>();
ConsumerRecord<String, String> record1 = new ConsumerRecord<>(topic, testPartition, startOffset, testKey1, testValue1);
Expand Down
Loading