[
https://issues.apache.org/jira/browse/KAFKA-1952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14321131#comment-14321131
]
Jun Rao commented on KAFKA-1952:
--------------------------------
The issue is that after purgatory refactoring, we make a checkSatisfied() call
on every key in the request. However, each checkSatisfied() goes through every
partition in the request. So, if you have 3K partitions in the request (like
the replica fetcher), we will be making 9 million calls to methods like
getPartition(). That's what's burning the CPU. This issue seems to only happen
in a cluster when there is no data being produced.
Attach a patch that addresses the issue.
> High CPU Usage in 0.8.2 release
> -------------------------------
>
> Key: KAFKA-1952
> URL: https://issues.apache.org/jira/browse/KAFKA-1952
> Project: Kafka
> Issue Type: Bug
> Affects Versions: 0.8.2.0
> Reporter: Jay Kreps
> Assignee: Jun Rao
> Priority: Critical
> Fix For: 0.8.2.0
>
> Attachments: kafka-1952.patch
>
>
> Brokers with high partition count see increased CPU usage when migrating from
> 0.8.1.1 to 0.8.2.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
