Hi, when you start your POC client do you rely on the default initial/maximal heap size of the VM or do you explicitly specify some values with the -Xms & -Xmx options? As the load you mentioned in your case could be effectively "high" I could imagine that specifiying those VM-bootstrap parameters could help.
IMHO that's a good idea to set them always to the same target value, that's instead of: -Xms512m -Xmx1024m One would better do: -Xms1024m -Xmx1024m As in the first case when the java native process tries to extend the heap size under high memory load, it could possibly not get it from the OS causing OOME as in your case (as the box is already under high load), however setting them to same value would make sure that *if* your process bootstraps successfully it has that required 1024m already in it's "pocket". I assume this thread also corresponds to the issue you reported at [1] [1] http://camel.465427.n5.nabble.com/Topic-consume-with-polling-connection-thread-Need-exmple-td5012908.html Babak -- View this message in context: http://camel.465427.n5.nabble.com/Camle-ActiveMQ-connection-polling-and-threading-issue-tp5012594p5013139.html Sent from the Camel - Users mailing list archive at Nabble.com.
