You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@activemq.apache.org by "Gary Tully (JIRA)" <ji...@apache.org> on 2016/09/01 15:29:20 UTC

[jira] [Updated] (AMQ-6413) kahadb concurrentStoreAndDispatch - producer audit managed by store is skipped

     [ https://issues.apache.org/jira/browse/AMQ-6413?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gary Tully updated AMQ-6413:
----------------------------
    Description: 
failover replay or inflight sends are trapped by the producer audit that is backed by the persistence adapter audit.
The intent is to drop any resend that has made it into the store.
With concurrentStoreAndDispatch messages can skip the store but also can skip the audit.
A failover resend can get delivered to the store and replayed to the cursor. 
AMQ-6406 and the message will go the the DLQ or get dispatched if the cursor audit is disabled/exhausted.

Pre AMQ-6406 the message would not get dispatched till restart, holding up gc.

  was:
failover replay or inflight sends are trapped by the producer audit that is backed by the persistence adapter audit.
The intent is to drop any resend that has made it into the store.
With concurrentStoreAndDispatch messages can skip the store but also can skip the audit.
A failover resend can get delivered to the store and replayed to the cursor. 
AMQ-6406 and the message will go the the DLQ or get dispatched if the cursor audit is disabled/exhausted.


> kahadb concurrentStoreAndDispatch - producer audit managed by store is skipped
> ------------------------------------------------------------------------------
>
>                 Key: AMQ-6413
>                 URL: https://issues.apache.org/jira/browse/AMQ-6413
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Message Store
>    Affects Versions: 5.14.0
>            Reporter: Gary Tully
>            Assignee: Gary Tully
>             Fix For: 5.15.0
>
>
> failover replay or inflight sends are trapped by the producer audit that is backed by the persistence adapter audit.
> The intent is to drop any resend that has made it into the store.
> With concurrentStoreAndDispatch messages can skip the store but also can skip the audit.
> A failover resend can get delivered to the store and replayed to the cursor. 
> AMQ-6406 and the message will go the the DLQ or get dispatched if the cursor audit is disabled/exhausted.
> Pre AMQ-6406 the message would not get dispatched till restart, holding up gc.



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