[jira] [Commented] (KAFKA-1120) Controller could miss a broker state change

["Jun Rao (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KAFKA-1120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13814096#comment-13814096 ] 

Jun Rao commented on KAFKA-1120:
--------------------------------

The root cause of this issue is that the controller missed a broker state change that it should have seen and therefore didn't make the correct decision. One way to fix that is for the controller to store the creation time of a broker registration. That way, on a broker change event, the controller can see if there has been any broker whose registration time has changed. We can then force a leader election on affected partitions accordingly. 

> Controller could miss a broker state change 
> --------------------------------------------
>
>                 Key: KAFKA-1120
>                 URL: https://issues.apache.org/jira/browse/KAFKA-1120
>             Project: Kafka
>          Issue Type: Bug
>          Components: core
>    Affects Versions: 0.8.1
>            Reporter: Jun Rao
>
> When the controller is in the middle of processing a task (e.g., preferred leader election, broker change), it holds a controller lock. During this time, a broker could have de-registered and re-registered itself in ZK. After the controller finishes processing the current task, it will start processing the logic in the broker change listener. However, it will see no broker change and therefore won't do anything to the restarted broker. This broker will be in a weird state since the controller doesn't inform it to become the leader of any partition. Yet, the cached metadata in other brokers could still list that broker as the leader for some partitions. Client requests routed to that broker will then get a TopicOrPartitionNotExistException. This broker will continue to be in this bad state until it's restarted again.



--
This message was sent by Atlassian JIRA
(v6.1#6144)