[jira] [Commented] (KAFKA-3042) updateIsr should stop after failed several times due to zkVersion issue

["Martin Nowak (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KAFKA-3042?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15618210#comment-15618210 ] 

Martin Nowak commented on KAFKA-3042:
-------------------------------------

Yes, still happens with 0.10.0.1. Also just ran into this after the ZooKeeper cluster wasn't available for a while.
The fact that the ZooKeeper cluster recovers itself, but Kafka instances (2 out of 3 in my case) remain unavailable until a manual restart, makes it fairly unreliable.
Not sure why you'd loose messages though, your client just shouldn't be able to write.
Yes, ZooKeeper setups should be improved to be reliable, but that's a misleading angle, because session expiration can happen for several reasons and Kafka needs to properly deal with it.

> updateIsr should stop after failed several times due to zkVersion issue
> -----------------------------------------------------------------------
>
>                 Key: KAFKA-3042
>                 URL: https://issues.apache.org/jira/browse/KAFKA-3042
>             Project: Kafka
>          Issue Type: Bug
>    Affects Versions: 0.8.2.1
>         Environment: jdk 1.7
> centos 6.4
>            Reporter: Jiahongchao
>             Fix For: 0.10.2.0
>
>         Attachments: controller.log, server.log.2016-03-23-01, state-change.log
>
>
> sometimes one broker may repeatly log
> "Cached zkVersion 54 not equal to that in zookeeper, skip updating ISR"
> I think this is because the broker consider itself as the leader in fact it's a follower.
> So after several failed tries, it need to find out who is the leader



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)