[ 
https://issues.apache.org/jira/browse/TOMEE-2027?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15928303#comment-15928303
 ] 

Bodo Pfelzer commented on TOMEE-2027:
-------------------------------------

You are right JobStoreTX makes it work. And 
org.apache.openejb.core.timer.TimerData#registerTimerDataSynchronization() 
really wants to use quartz outside any transaction. I think this is closely 
related to my pull request

https://github.com/apache/tomee/pull/64

which asks, who is responsible for storing persistent timers. Perhaps 
registerTimerDataSynchronization() should only be called, if JobStoreCMT is not 
used. In the JobStoreCMT case EbicsTimerServiceImpl should pass it directly to 
quartz.

In fact I only want to use persistent timers, which are part of the jee spec. 
It is hard to google, how it should be used in TomEE. I only got links with 
JobStoreCMT, and even QuartzPersistenceForEjbTimersTest in openejb-core uses 
it. A consolidated documentation would be fine.

> Persistent timers eat connection pool
> -------------------------------------
>
>                 Key: TOMEE-2027
>                 URL: https://issues.apache.org/jira/browse/TOMEE-2027
>             Project: TomEE
>          Issue Type: Bug
>          Components: TomEE Core Server
>    Affects Versions: 7.0.4
>            Reporter: Bodo Pfelzer
>         Attachments: catalina.out, QuartzPersistenceForEJBTimers2Test.java, 
> SimpleBean.java, SimpleLocal.java, system.properties
>
>
> I use persistent timers with a configuration similar to the one use in 
> openejb-core test QuartzPersistenceForEJBTimersTest (see attached 
> system.properties). A simple bean uses this timer service to create new a new 
> timer whenever one expires. After 33 iterations the underlying connection 
> pool is exhausted (see catalina.out). The same example works in tomee-1.7.x.
> I also tried to create an equivalent maven test based by modifying 
> QuartzPersistenceForEJBTimersTest. But I was not able to reproduce the 
> problem there. But it seems, that a different connection pool is used in the 
> test environment.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to