accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Eric Newton (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-3117) UnorderedWorkAssignerReplicationIT failing in nightly tests
Date Fri, 12 Sep 2014 21:04:34 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-3117?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14132108#comment-14132108
] 

Eric Newton commented on ACCUMULO-3117:
---------------------------------------

I modified the way that the mutation queue works.  It used to flush when any connection had
more than 1M of mutations (by default) queued up for the WAL.
Now it keeps a global count of all the mutations queued up in the entire server, but I made
the limit higher (50M).

Simply changing this new limit back to 1M causes the tests to succeed.  Why it should be so,
I don't yet know. [~elserj] any ideas?


> UnorderedWorkAssignerReplicationIT failing in nightly tests
> -----------------------------------------------------------
>
>                 Key: ACCUMULO-3117
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-3117
>             Project: Accumulo
>          Issue Type: Bug
>          Components: test
>            Reporter: Eric Newton
>            Assignee: Eric Newton
>             Fix For: 1.7.0
>
>
> UnorderedWorkAssignerReplicationIT#dataReplicatedToThePeer fails:
> "Drain did not finish within 30 seconds"
> UnorderedWorkAssignerReplicationIT#dataWasReplicatedToThePeerWithoutDrain fails:
> "test timed out after 300000 milliseconds"



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message