[Vagnekman-users] Batch job processing at Ekman

Michael Schliephake michs at kth.se
Fri May 13 12:03:23 CEST 2011


Dear Users,

The batch processing has been started 11.15 again.

The evaluation of the hardware showed no special events. A few nodes showed hardware failures as usual.

A possible explanation is that caused by these hardware failures other nodes in the parallel jobs hung and could not be restarted as they should. The high number of affected nodes could be a coincidence from a higher number of jobs ending more or less at the same time. 


Best regards -- Med vänlig hälsning

Michael




More information about the Vagnekman-users mailing list