kafka-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ewen Cheslack-Postava (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (KAFKA-2477) Replicas spuriously deleting all segments in partition
Date Tue, 13 Oct 2015 06:57:05 GMT

    [ https://issues.apache.org/jira/browse/KAFKA-2477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14954511#comment-14954511
] 

Ewen Cheslack-Postava commented on KAFKA-2477:
----------------------------------------------

[~cpsoman] Beyond applying to 0.8.2.0 with the patch, any reason not to update to 0.8.2.2
and apply the patch on top of that, where it definitely applies cleanly? It looks like 8 patches,
and some of the patches on top of 0.8.2.0 are likely to be useful if you might have a large
number of partitions or use snappy compression, among other key fixes. Maybe you're not hitting
any of the critical fixes in those releases, but since they're low risk maybe catching up
with the latest release and only having a minor patch would simplify things?

> 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: HÃ¥kon Hitland
>            Assignee: Jiangjie Qin
>             Fix For: 0.9.0.0
>
>         Attachments: Screen Shot 2015-10-10 at 6.54.44 PM.png, kafka_log.txt, kafka_log_trace.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)

Mime
View raw message