Paul, Yes that should work. This is based on the concurrent.jar that gets a permit
waiting for the timeout second u provide in the -ds.xml
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3833800#3833800
Reply to the post :
http://www.jboss.org/index.html?modul
Controlling the number of concurrent requests.
It uses strict pooling. This should be combined with the
| http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3830080#3830080
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3830080
---
Also what are the benefit (and drawbacks) of changing the max-pool-size setting.
Regards,
Stephane
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3829846#3829846
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3829846
Hey Adrian,
So you suggest to open/close the connection every time I need access to the queue.
Well ok but could it be the cause of my problem?
I don't get you with this pool size settings. My scenario is:
RequresNew - New transaction is created
- Create session
- Create receiver
- Read message
The default pool size is 20
There is an *implicit*
| 20
|
in jms-ds.xml
The interrupt is a transaction timeout.
Allocating a connection for the lifetime of the bean is not a very good use of
resources.
The things are pooled for gods sake, just open and close them as you need them,
they don
I've been seeing this, having done some recent performance testing on JBoss 3.2.3. I
suspect this is a JBoss bug. This happens after about 50,000 message processed by a
MDB which is using the java:/JmsXA connection factory (transacted.)
I'm not sure why the connection pool would block; why wo
I forgot to mention that we use java:/JmsXA as connection factory.
Regards,
Stephane
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3829812#3829812
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3829812
--
