Hi David, thank you for your reply, I remember your precious help on the Connector mailing list.
David Jencks wrote:
you should expect a clustered solution to have slightly more latency than an equivalently configured (with same commit option) non clustered solution: there is more communication. However by adding more servers you should be able to get higher throughput as well (again, with the same commit option).
Of course, Thing is that I need to use option B on cluster as stated on Jboss docs. In this case I loose completely the caching mechanism. This is the reason because I'm seeing performances degradation.
I looked at the seppuku pattern but it's tied to BEA features.
It's very interesting the ACE pattern that it's vendor independent
http://www.theserverside.com//patterns/thread.jsp?thread_id=10610&article_count=47#35119
But also this solution implies a large code rewriting an right now I'm not able to completely evaluate it.
I am not an expert on this!!
I know, I hope to receive some suggestion from Sacha
My understanding is that guaranteed correct reads can only be obtained with a distributed cache which is now under development.
Any planned date ;-) ?
If I do not use entity beans. Is there any way to have a singleton Hashtable available to my application ?
Thank you in advance
Davide
-------------------------------------------------------
This SF.net email is sponsored by: Etnus, makers of TotalView, The debugger for complex code. Debugging C/C++ programs can leave you feeling lost and disoriented. TotalView can help you find your way. Available on major UNIX and Linux platforms. Try it free. www.etnus.com
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-user
