Troubleshoot Kafka

Search by the error you see. Each page says what it means, the likely causes in order, and how to confirm and fix each one.

Or start from what you see

Area

Most common

TimeoutException: Expiring 5 record(s) for orders-0:120000 ms has passed since batch creationThe producer accepted the records, but the batch was not delivered within delivery.timeout.ms (120000 ms, two minutes, by default). The client then failed every record in the batch with a TimeoutException. The number in the message is how many records were in the batch, and the time is how long ago the batch was created.ProducerHas a drill
NOT_ENOUGH_REPLICAS (NotEnoughReplicasException)A producer using acks=all wrote to a partition whose in-sync replica set has fewer members than min.insync.replicas. The protocol's error table describes it as "Messages are rejected since there are fewer in-sync replicas than required." NOT_ENOUGH_REPLICAS_AFTER_APPEND is the case where the low ISR size was discovered after the message was already appended to the log. Its Javadoc notes that producer retries will cause duplicates.ProducerHas a drill
Connection to node -1 (localhost/127.0.0.1:9092) could not be established. Node may not be available.Before a client has fetched metadata, it knows only the addresses in bootstrap.servers, and the client gives those addresses negative node ids (-1, -2 and so on). So a message naming node -1 means a connection to a bootstrap address could not be opened. Brokers get their real ids from the metadata the client fetches after that first connection.Connecting
CommitFailedException: Commit cannot be completed since the group has already rebalanced and assigned the partitions to another memberThe consumer tried to commit offsets after the group had already rebalanced and moved its partitions to another member. The Javadoc says this can happen when a group rebalance completes before the commit could be applied, and that the commit cannot generally be retried because some partitions may already be assigned to another member. The full message adds that the time between calls to poll() was longer than max.poll.interval.ms, which typically implies the poll loop is spending too much time on message processing.ConsumerHas a drill
Consumer group keeps rebalancingWhenever membership changes, the group rebalances. With an eager protocol a consumer must revoke all its partitions before taking part (the RebalanceProtocol Javadoc), and KIP-848 describes the classic protocol as relying on a group-wide synchronization barrier. A rebalance storm is when this happens again and again, so consumers spend their time rejoining instead of consuming.ConsumerHas a drill
Under-replicated partitions greater than zeroA partition is under-replicated when its in-sync replica set has fewer members than its replica list. The monitoring page defines the UnderReplicatedPartitions metric as the number of partitions with |ISR| < |all replicas|, and its normal value is 0. The data is still served by the leader, but you have less redundancy than you configured.ClusterHas a drill

All errors

Showing 12 of 12