[
https://issues.apache.org/jira/browse/ISIS-1097?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14365103#comment-14365103
]
Oscar Bou commented on ISIS-1097:
---------------------------------
In our case we've discovered that the long-time running queries we've detected
are due to a infinite recursion action invocation.
But that code is heavily dependant on the EventBusService. So we're suspecting
about the queueing behavior we notified on the user list, and whose solution
might be the one proposed at [1] (implementing one based on Axon), or create a
new Guava's Event Bus Service descendant that instantiates a dispatcher that
does not enqueue calls.
[1] https://issues.apache.org/jira/browse/ISIS-1028
> Suspicion that Isis may be leaving DB connections open under certain
> circumstances, leading to instability
> ----------------------------------------------------------------------------------------------------------
>
> Key: ISIS-1097
> URL: https://issues.apache.org/jira/browse/ISIS-1097
> Project: Isis
> Issue Type: Bug
> Components: Core
> Affects Versions: core-1.8.0
> Reporter: Dan Haywood
> Assignee: Dan Haywood
> Fix For: 1.9.0
>
>
> Report from the field:
> Users reporting that the app was slow. We've seen some active sessions in
> Tomcat that didn't correspond to any user, but also 3 database connections on
> PostgreSQL that were "IN TRANSACTION" for a long time.
> Killing the Tomcat and database sessions the correct speed has been recovered
> ~~~~
> One monitoring option to consider (open source with commercial support) is
> hyperic,
> * http://sourceforge.net/projects/hyperic-hq/
> *
> https://my.vmware.com/web/vmware/details?downloadGroup=VRHQ_584_OSS&productId=378
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
