Hi,
Around 50% of jobs going to roma fail - stdout and stderr are empty.
Those which run successfully always run on hosts
t2-wn-02.roma1.infn.it or t2-wn-03.roma1.infn.it
but I see that there are hosts 01-11. Could there be some problem with
the pbs/ssh on some of the nodes?
Also there are 3 sites where
GlueCEStateFreeCPUs+GlueCEStateRunningJobs != GlueCEInfoTotalCPUs
t2-ce-01.lnl.infn.it, t2-ce-01.mi.infn.it, t2-ce-01.roma1.infn.it
and 3 where
GlueCEStateWaitingJobs>0 &&GlueCEStateRunningJobs ==0
grid003.ft.uam.es, gw39.hep.ph.ic.ac.uk,t2-ce-01.roma1.infn.it
But hey, who's counting.
Cheers,
Rod.
|