[jira] [Updated] (KAFKA-1355) Reduce/optimize update metadata requests sent during leader election

["Joel Koshy (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/KAFKA-1355?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Joel Koshy updated KAFKA-1355:
------------------------------

    Attachment: KAFKA-1355_2014-04-04_13:48:34.patch

> Reduce/optimize update metadata requests sent during leader election
> --------------------------------------------------------------------
>
>                 Key: KAFKA-1355
>                 URL: https://issues.apache.org/jira/browse/KAFKA-1355
>             Project: Kafka
>          Issue Type: Bug
>    Affects Versions: 0.8.1
>            Reporter: Joel Koshy
>            Priority: Critical
>             Fix For: 0.8.1.1
>
>         Attachments: KAFKA-1355_2014-04-04_13:48:34.patch
>
>
> This is part of the investigation into slow shutdowns in 0.8.1. While
> logging contributes to bulk of the regression, this one also adds
> quite a bit of overhead:
> In addLeaderAndIsrRequest (called for every partition that is led by the
> broker being shut down) we also add an UpdateMetadataRequest - each call to
> addUpdateMetadataRequests does two traversals over _all_ (global)
> partitions. I think it should be straightforward to optimize this a bit.
> Marking as critical, since it is not as big an overhead as the logging.



--
This message was sent by Atlassian JIRA
(v6.2#6252)