The problem with going up to more than say 5 or 6 nodes is that replication and
2 phase commits across a cluster become a big overhead and you will find that
the app does not scale. I'm assuming you're using synchronous replication.
As a solution, to help things scale better, either use asunchr
Hi everyone,
JBosscache seems to be a great cache but our tests did not demonstrate the
expectected results. We have only 8 nodes so far and we would like to go up to
100nodes. With 8 nodes, we already experienced lock problems in both modes
(optimistic, pessimistic).
So few questions come
more informations about my problem:
We still experience concurrent problem whether the cache is configured
optimistic or pessimistic. Our test is simple:
1- A POJO containing a collection (HashMap) is written into the cache by
JVM1 (line 5-6).
2- Then JVM1 to JVM8 are putting objects
So among 8 node cluster, only one node is doing write with single-thread write
while others are reading?
View the original post :
http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4016163#4016163
Reply to the post :
http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4016
