Xavier Lange created KAFKA-4412: ----------------------------------- Summary: Replication fetch stuck in loop on offset null Key: KAFKA-4412 URL: https://issues.apache.org/jira/browse/KAFKA-4412 Project: Kafka Issue Type: Bug Components: replication Reporter: Xavier Lange
I kicked off a cluster rebalance and it never completed. I had to look at node eth0 traffic to see there was a constant 60MB/s (I'm usually at about 5MB/s ingest). The /kafka/logs/server.log was looping like this, then I deleted the topic in question to make it stop: {quote} [2016-11-15 18:21:27,745] ERROR Found invalid messages during fetch for partition [cisco-2016.11.13,19] offset 861323 error null (kafka.server.ReplicaFetcherThread) [2016-11-15 18:21:27,755] ERROR Found invalid messages during fetch for partition [cisco-2016.11.13,19] offset 861323 error null (kafka.server.ReplicaFetcherThread) [2016-11-15 18:21:27,773] ERROR Found invalid messages during fetch for partition [cisco-2016.11.13,19] offset 861323 error null (kafka.server.ReplicaFetcherThread) [2016-11-15 18:21:27,788] ERROR Found invalid messages during fetch for partition [cisco-2016.11.13,19] offset 861323 error null (kafka.server.ReplicaFetcherThread) [2016-11-15 18:21:27,847] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,19] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,852] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,11] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,853] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,9] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,855] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,3] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,856] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,16] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,857] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,2] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:27,858] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,19] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,012] INFO Deleting index /data/cisco-2016.11.13-19/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,016] INFO Deleted log for partition [cisco-2016.11.13,19] in /data/cisco-2016.11.13-19. (kafka.log.LogManager) [2016-11-15 18:21:28,024] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,11] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,165] INFO Deleting index /data/cisco-2016.11.13-11/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,165] INFO Deleted log for partition [cisco-2016.11.13,11] in /data/cisco-2016.11.13-11. (kafka.log.LogManager) [2016-11-15 18:21:28,167] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,9] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,232] INFO Deleting index /data/cisco-2016.11.13-9/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,232] INFO Deleted log for partition [cisco-2016.11.13,9] in /data/cisco-2016.11.13-9. (kafka.log.LogManager) [2016-11-15 18:21:28,242] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,3] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,341] INFO Deleting index /data/cisco-2016.11.13-3/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,342] INFO Deleted log for partition [cisco-2016.11.13,3] in /data/cisco-2016.11.13-3. (kafka.log.LogManager) [2016-11-15 18:21:28,375] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,16] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,465] INFO Deleting index /data/cisco-2016.11.13-16/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,466] INFO Deleted log for partition [cisco-2016.11.13,16] in /data/cisco-2016.11.13-16. (kafka.log.LogManager) [2016-11-15 18:21:28,469] INFO [ReplicaFetcherManager on broker 0] Removed fetcher for partitions [cisco-2016.11.13,2] (kafka.server.ReplicaFetcherManager) [2016-11-15 18:21:28,486] INFO Deleting index /data/cisco-2016.11.13-2/00000000000000000000.index (kafka.log.OffsetIndex) [2016-11-15 18:21:28,486] INFO Deleted log for partition [cisco-2016.11.13,2] in /data/cisco-2016.11.13-2. (kafka.log.LogManager) {quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332)