From: DimitryASuplatov (genesup_at_gmail.com)
Date: Sun Feb 07 2010 - 04:39:40 CST
sometimes I get errors like these.
In all my cases this is an indication of hardware problems on the
cluster. Try writing problem report to your cluster administration
specifying the content of machinefile that contains nodes ids that
were used to run the task. The only difference with your case is that
I usually run on 64-128 CPUs thus correct hardware installation is
more essential. In your case 4CPUs should be located on one node
nevertheless problems are still possible.
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:22:42 CST