[
https://issues.apache.org/jira/browse/AMQ-3196?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Gary Tully resolved AMQ-3196.
-----------------------------
Resolution: Fixed
fix in http://svn.apache.org/viewvc?rev=1076278&view=rev
two less calls to underlying store for durable subs reactivation. Initial
response is now quite nippy for a large backlog with the jdbc store.
> Speed up initial message delivery for offline durable sub with
> keepDurableSubsActive=true and JDBC store
> --------------------------------------------------------------------------------------------------------
>
> Key: AMQ-3196
> URL: https://issues.apache.org/jira/browse/AMQ-3196
> Project: ActiveMQ
> Issue Type: Improvement
> Components: Message Store
> Affects Versions: 5.4.2
> Reporter: Gary Tully
> Assignee: Gary Tully
> Labels: DurableSubscriptions, JDBC, Offline, Performance
> Fix For: 5.5.0
>
>
> Follow on from https://issues.apache.org/jira/browse/AMQ-3188
> Durablesubs are kept active by default, but their cursors are stopped
> requiring a read of the store to determine the message backlog. With a large
> prioritised backlog and th JDBC store, getting the store size can be costly
> as it requires a full table scan.
> Currently there are three calls to getStoreSize, there should be only one
> when the sub is initially created.
> The implication is that on reactivation of an offline durable sub, there is a
> delay of an additional potentially slow jdbc call before dispatch resumes.
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
