kafka-jira mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jan Filipiak (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (KAFKA-6056) LogCleaner always cleaning into 1 Segment per sizegroup might exeed relativ offset range
Date Thu, 12 Oct 2017 05:18:00 GMT

    [ https://issues.apache.org/jira/browse/KAFKA-6056?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16201455#comment-16201455
] 

Jan Filipiak commented on KAFKA-6056:
-------------------------------------

A valid workaround for partitions containing the segments should be to replicate them to a
broker of a version >10.2 wich will then roll the segment while building up the replica.
As there is no problem in the code actually. One could think of including this into the upgrade
notices? 

> LogCleaner always cleaning into 1 Segment per sizegroup might exeed relativ offset range
> ----------------------------------------------------------------------------------------
>
>                 Key: KAFKA-6056
>                 URL: https://issues.apache.org/jira/browse/KAFKA-6056
>             Project: Kafka
>          Issue Type: Bug
>          Components: core, log
>    Affects Versions: 0.11.0.0
>            Reporter: Jan Filipiak
>            Priority: Minor
>
> After having an Issue with compaction stopping for some time. It can be an issue that
the LogCleaner will always clean into 1 Segment per sizegroup. 
> Usually  the Log enforces a maximum distance between min and max offset in a LogSegment.
If that Distance would be exeeded in maybeRoll() a new logsegment would be rolled. I assume
this is because relative offset might be stored as integer. The LogCleaner OTOH is not going
to roll a new LogSegment as its only ever using 1 Segment to clean into. 
> A lenghty discussion about this can be found in the slack community:
> https://confluentcommunity.slack.com/archives/C49R61XMM/p1506914441000005
> The observed stacktrace is as follows:
> https://gist.github.com/brettrann/ce52343692696a45d5b9f4df723bcd14
> I could imagin also enfocing Integer.MAX_VALUE as offset distance in
> groupSegmentsBySize in the LogCleaner to make sure a Segment doesnt exeed this limit.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message