restart error in cluster

Date view Thread view Subject view Author view Attachment view

From: fengguang tian (fernyabc_at_gmail_dot_com)
Date: Fri Apr 02 2010 - 09:50:11 PDT

  • Next message: Balazs Gerofi: "BLCR @ BG/P?" 
    Hi,
when in a cluster with two machines, the checkpoint and restart works fine,
but when the cluster has three or more machines, when restart, some error
happened:

mpirun noticed that process rank 6 with PID 27878 on node xxxx exited on
signal 11 (Segmentation fault).

some segmentation fault happened. I am using NFS file during the machines.

cheers
  • Next message: Balazs Gerofi: "BLCR @ BG/P?"
    Date view Thread view Subject view Author view Attachment view