[jira] [Commented] (HELIX-472) Errors should be cleaned up less frequently

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

    [ https://issues.apache.org/jira/browse/HELIX-472?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14064043#comment-14064043 ] 

Hudson commented on HELIX-472:
------------------------------

UNSTABLE: Integrated in helix #1264 (See [https://builds.apache.org/job/helix/1264/])
[HELIX-472] Errors should be cleaned up less frequently (kanak: rev f952cb7606f6de91dbba4f5e4597e45259ec5bf6)
* helix-core/src/test/java/org/apache/helix/integration/TestSchedulerMessage.java
* helix-core/src/main/java/org/apache/helix/manager/zk/ZKHelixManager.java
* helix-core/src/main/java/org/apache/helix/monitoring/ZKPathDataDumpTask.java
* helix-core/src/test/java/org/apache/helix/monitoring/TestZKPathDataDumpTask.java


> Errors should be cleaned up less frequently
> -------------------------------------------
>
>                 Key: HELIX-472
>                 URL: https://issues.apache.org/jira/browse/HELIX-472
>             Project: Apache Helix
>          Issue Type: Bug
>            Reporter: Kanak Biscuitwala
>            Assignee: Kanak Biscuitwala
>
> As part of HELIX-452, we decreased the status update cleanup window to 15 minutes and the error cleanup window to 45 minutes. This makes it more troublesome trying to look back and find historical errors, and these are proportional to numPartitions * numSessionsInWindow number of znodes (much less than number of possible status updates). We should increase the error cleanup threshold to a day.



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