-
1. Re: OutOfMemoryError
Christian Riege Feb 21, 2002 5:17 AM (in response to Khoa Nguyen)there is one which has been fixed in current 2.4 CVS and will be incorporated into 2.4.5. This is related to EJB locking, see SourceForge bug #504895 for more information.
i don't know if this is biting you though as the leak is rather slow. -
2. Re: OutOfMemoryError
Hiram Chirino Feb 22, 2002 11:21 PM (in response to Khoa Nguyen)Does that version have the message cache feature?? If it doesn't then all those messages allways say in memory until they are consumed.
With the message cache feature, JBossMQ pages messages out to disk if is running low on memory.
Regards,
Hiram -
3. Re: OutOfMemoryError
Khoa Nguyen Feb 25, 2002 7:25 PM (in response to Khoa Nguyen)is message caching in 2.4.4, if it is where is it configured?
-
4. Re: OutOfMemoryError
Hiram Chirino Feb 26, 2002 1:42 AM (in response to Khoa Nguyen)I just checked... It is not in 2.4.4
It is a 3.0 only feature. -
5. Re: OutOfMemoryError
Peter Antman Feb 26, 2002 3:30 PM (in response to Khoa Nguyen)Untill a finish the backport ;-)
//Peter -
6. Re: OutOfMemoryError
Khoa Nguyen Mar 1, 2002 1:39 PM (in response to Khoa Nguyen)Please let me know when the backport is available for 2.4.4. I can't use 3.0 for a production application because it's still in Beta.
- Khoa -
7. Re: OutOfMemoryError
Peter Antman Mar 1, 2002 4:55 PM (in response to Khoa Nguyen)Keep an eay on the forum...
//Peter -
8. Re: OutOfMemoryError
samlam Mar 5, 2002 2:12 AM (in response to Khoa Nguyen)Anyone please let me know that the memory leak fault of
JMS/Joss has been fixed or still under investigation.
Is 24x7 avaiable on joss/jms
many forum in serval jms provider that the stability
of jms protocal still un stable for the time being. -
9. Re: OutOfMemoryError
Khoa Nguyen Mar 5, 2002 5:03 PM (in response to Khoa Nguyen)All my messages are Durable and are consumed by the MDB immediately after receipt. I've only allocated a maximum 1 MDB. The messages shouldn't be in memory very long so I'm not convinced the message caching will fix the problem. Could the memory leak be somewhere else?
Here's a snipet of the architecture
- stateless session bean put message in queue
- MDB 1 consumes message in queue
- MDB 1 creates a worker (ie. stateless session bean), process the message, put message in another queue
- MDB 2 consumes message
- done
- Khoa