[fix][io] Kafka Source connector maybe stuck

[shibd]

Motivation

The current implementation of Kafka source connector, that KafkaRecord does not implement the fail() method.

pulsar/pulsar-io/kafka/src/main/java/org/apache/pulsar/io/kafka/KafkaAbstractSource.java

Line 222 in bbae607

protected static class KafkaRecord<V> implements Record<V> {

If PulsarSink send a message to pulsar failed, will call record.fail().

pulsar/pulsar-functions/instance/src/main/java/org/apache/pulsar/functions/sink/PulsarSink.java

Lines 194 to 196 in 43a9898

	if (failSource) {
	srcRecord.fail();
	}

But because KafkaRecord does not implement it, this futures will never end.

pulsar/pulsar-io/kafka/src/main/java/org/apache/pulsar/io/kafka/KafkaAbstractSource.java

Line 178 in 495b141

CompletableFuture.allOf(futures).get();

This causes the Kafka consumer to leave the consumer group due to more than max.poll.interval.ms.

2024-04-11T14:06:25,730-0400 [kafka-coordinator-heartbeat-thread | kafka2pulsarConsumer] INFO  org.apache.kafka.clients.consumer.internals.AbstractCoordinator - [Consumer clientId=consumer-1, groupId=consumer-g-1] Member consumer-1 sending LeaveGroup request to coordinator localhost:9092 (id: 2147483644 rack: null) due to consumer poll timeout has expired. This means the time between subsequent calls to poll() was longer than the configured max.poll.interval.ms, which typically implies that the poll loop is spending too much time processing messages. You can address this either by increasing max.poll.interval.ms or by reducing the maximum size of batches returned in poll() with max.poll.records.

Modifications

• Implement fail() method for KakfaRecord.
• Let CompletableFuture.allOf(futures).get() support a timeout, set to 2/3 of max.poll.interval.ms. This adjustment is to ensure the interval between two poll requests does not exceed Kafka max.poll.interval.ms.
  • When it times out, it will trigger a connector failure, thereby restarting the connector.

Verifying this change

• Add throwExceptionBySendFail and throwExceptionBySendTimeOut to cover it.

Documentation

• doc
• doc-required
• doc-not-needed
• doc-complete

Matching PR in forked repository

PR in forked repository:

[github-actions]

@shibd Please add the following content to your PR description and select a checkbox:

- [ ] `doc` <!-- Your PR contains doc changes -->
- [ ] `doc-required` <!-- Your PR changes impact docs and you will update later -->
- [ ] `doc-not-needed` <!-- Your PR changes do not impact docs -->
- [ ] `doc-complete` <!-- Docs have been already added -->

[lhotari]

LGTM. Good work @shibd

[Demogorgon314]

LGTM

[nicoloboschi]

LGTM