parallel jobs with 3.22

Yacine

At the moment I am tryng to do few tests on two clusters we have here: One with mosix operating system ( Linux 2.4.26-mosix Red Hat 8.0). The second with the SGE N1GE 6.0 batch environment. I am having few problems: In both clusters I am getting this error: "PNP: Starting TRACKER task on "cla00" and "cla01" for monitoring host and process failures". These nodes correspond to the first and second node in the parallel job

This seems (but i am not very sure)to be causing the job to stop after some time in the cluster with the batch environment, while the runs don't seem to be affected on the other cluster.

It would be very helpfull if someone can give me some guidance on this matter