You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@qpid.apache.org by "ASF subversion and git services (JIRA)" <ji...@apache.org> on 2014/04/07 17:40:17 UTC

[jira] [Commented] (QPID-5513) HA backup fails if number of replicated queues exceeds number of channels.

    [ https://issues.apache.org/jira/browse/QPID-5513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13961936#comment-13961936 ] 

ASF subversion and git services commented on QPID-5513:
-------------------------------------------------------

Commit 1585507 from aconway@apache.org in branch 'qpid/trunk'
[ https://svn.apache.org/r1585507 ]

QPID-5666: HA fails with resource-limit-exceeded: Exceeded replicated queue limit

This is regression introduced in r1561206: CommitDate: Fri Jan 24 21:54:59 2014 +0000
  QPID-5513: HA backup fails if number of replicated queues exceeds number of channels.

Fixed by the current commit. PrimaryQueueLimits was not taking account of queues already
on the broker prior to promotion.

> HA backup fails if number of replicated queues exceeds number of channels.
> --------------------------------------------------------------------------
>
>                 Key: QPID-5513
>                 URL: https://issues.apache.org/jira/browse/QPID-5513
>             Project: Qpid
>          Issue Type: Bug
>          Components: C++ Clustering
>    Affects Versions: 0.25
>            Reporter: Alan Conway
>            Assignee: Alan Conway
>             Fix For: 0.27
>
>
> HA uses a single link and a session per queue to replicate queue data. That means the number of replicated queues is limited by the number of channels, which is 32k in current Qpid.
> To reproduce:
> - create a 2 node cluster.
> - declare 32769 queues
> Actual Result:
> - no error observed by the client declaring the queues or in the primary log.
> - backup broker exits with a critical error saying "not enough channels"
> In the short term it is not straightforward to remove the limit, however the error reporting needs to be fixed:
> Desired result:
> - client declaring the queues receives a "resource-limit-exceeded" exception with appropriate message.
> - primary logs an error indicating we exceeded the replicated queue limit
> - primary and backup broker continue to function normally.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@qpid.apache.org
For additional commands, e-mail: dev-help@qpid.apache.org