[ https://issues.apache.org/jira/browse/KAFKA-2477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14730713#comment-14730713 ]
Håkon Hitland commented on KAFKA-2477: -------------------------------------- I don't see any out of sequence offsets. Here are a couple of recent examples. If I run with --deep-iteration, all offsets are present and sequential. The result on the replica is identical to the leader. --- [2015-09-02 23:43:03,379] ERROR [Replica Manager on Broker 0]: Error when processing fetch request for partition [log.event,3] offset 10591627212 from follower with correlation id 391785394. Possible cause: Request for offset 10591627212 but we only have log segments in the range 10444248800 to 10591627211. (kafka.server.ReplicaManager) offset: 10591627210 position: 994954613 isvalid: true payloadsize: 674 magic: 0 compresscodec: SnappyCompressionCodec crc: 4144791071 offset: 10591627211 position: 994955313 isvalid: true payloadsize: 1255 magic: 0 compresscodec: SnappyCompressionCodec crc: 1011806998 offset: 10591627213 position: 994956594 isvalid: true payloadsize: 1460 magic: 0 compresscodec: SnappyCompressionCodec crc: 4145284502 offset: 10591627215 position: 994958080 isvalid: true payloadsize: 1719 magic: 0 compresscodec: SnappyCompressionCodec crc: 444418110 ---- [2015-09-03 11:44:02,483] ERROR [Replica Manager on Broker 3]: Error when processing fetch request for partition [log.count,5] offset 69746066284 from follower with correlation id 239821628. Possible cause: Request for offset 69746066284 but we only have log segments in the range 68788206610 to 69746066280. (kafka.server.ReplicaManager) offset: 69746066278 position: 464897345 isvalid: true payloadsize: 674 magic: 0 compresscodec: SnappyCompressionCodec crc: 3013732329 offset: 69746066279 position: 464898045 isvalid: true payloadsize: 234 magic: 0 compresscodec: SnappyCompressionCodec crc: 3286064200 offset: 69746066283 position: 464898305 isvalid: true payloadsize: 486 magic: 0 compresscodec: SnappyCompressionCodec crc: 747917524 offset: 69746066285 position: 464898817 isvalid: true payloadsize: 342 magic: 0 compresscodec: SnappyCompressionCodec crc: 4283754786 offset: 69746066286 position: 464899185 isvalid: true payloadsize: 233 magic: 0 compresscodec: SnappyCompressionCodec crc: 2129213572 > Replicas spuriously deleting all segments in partition > ------------------------------------------------------ > > Key: KAFKA-2477 > URL: https://issues.apache.org/jira/browse/KAFKA-2477 > Project: Kafka > Issue Type: Bug > Affects Versions: 0.8.2.1 > Reporter: Håkon Hitland > Attachments: kafka_log.txt > > > We're seeing some strange behaviour in brokers: a replica will sometimes > schedule all segments in a partition for deletion, and then immediately start > replicating them back, triggering our check for under-replicating topics. > This happens on average a couple of times a week, for different brokers and > topics. > We have per-topic retention.ms and retention.bytes configuration, the topics > where we've seen this happen are hitting the size limit. -- This message was sent by Atlassian JIRA (v6.3.4#6332)