Hi everybody,
i'm newbie about clustering JBoss.
I have done a software which runs on jboss and now I would like to cluster it.
I know it's quite easy to do that, but let me explain deeper what exactly does 
it.
In few words it just runs a method which will make a lot of math 
calculation.This calculation will not gain to be parallelized so it's ok it 
runs on a singol core (no need of gridgain or similar).
Now it runs on a quadcore CPU, If I run once it takes 25% of total CPU (25% for 
each core) If I run 4 method the CPU goes to 100%. I see that if I run 6-8 
method, the resolution time doesn't increase a lot, anyway this is another 
problem.
I read jboss cluster is for a businness continuity scenario, that is good for 
me, but also to load balance a high number of requests, that is not my problem.
If I create a cluster, as example 2 node each quad core, and I run 2 method, 
which is the police of the cluster? I mean one node 25% and the other 25%?
I would like to fix the maximum number of optimization run at the same time, 
based on number of cores, in the above example let's say 8.
Will it balance the method automatically based upon cpu usage? Or it don't 
care? 
Can I start method manully on a singol defined cluster?

I'm sorry about my question, I know it probably seems a little bit nebulous, I 
need just a good advice where to start (now of course reading Cluster 
documentation)

thanks a lot 

nibe


View the original post : 
http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4184647#4184647

Reply to the post : 
http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4184647
_______________________________________________
jboss-user mailing list
[email protected]
https://lists.jboss.org/mailman/listinfo/jboss-user

Reply via email to