Re: [ERROR] NAKACK.handleXmitReq()

fc-fast-minimalthreads.xml. The one that I attached should be fine too: 
I started the test with tmp.xml on 3 boxes in our Atlanta labs this 
morning, and so far we're at 1.27 billion messages (total of 6 billion, 
2B/sender), and everything looks fine

Gianluca Puggelli wrote:
> Thanks Bela for the very quick answer.
>
> At the moment my primary target is to check the jgroups stability over
> a long time execution.
>
> On Friday 15 September 2006 10:00, you wrote:
>   
>> Those might be old retransmit messages, which are received after the 
>> messages have already been purged. To avoid this, change
>> - NAKACK.retransmit_timeout: remove 100 and 200: 
>> retransmit_timeout="300,600,1200,2400,4800"
>> - STABLE.max_bytes: set it to 2000000 from 400000
>> - I would also suggest to enable bundling in TCP_NIO: 
>> enable_bundling="true", this speeds up things
>>
>> BTW: I get much better results with tcp.xml right now, need to see why
>>
>> I assume you still receive all messages, correct ?
>>
>>     
>
> I don't know, because first attempt stopped after 34 hours with OutOfMemory 
> exception, while the second got OutOfMemory exception only after few hours.
>
> Which is the most suitable protocol for long time execution ?
>
> thanks and regards
>       gianluca 
>
>   

-- 
Bela Ban
Lead JGroups / Manager JBoss Clustering Group
JBoss - a division of Red Hat


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642