python program killed in pbs without info

I'm running a simple Python program using rpy2 in a cluster managed by pbs_torque. My problem is that the same piece of code works perfectly and other times is killed without logical reason or error trace. At random points during the execution of the program, the execution finishes. The only output error I get is this:

kill: 20: No such process

I do not need to mention, that I'm not printing that message. Some output error files even contain this message several 开发者_StackOverflowtimes. I've done some search but I have not found any relevant help. All the nodes of the cluster have the same configuration, and are supposed to run one instance of the program per core.

Suggestions are welcome.

Thanks