 |
 |
Hi Andreas, Andreas Haupt wrote: > On Tue, 2009-03-31 at 14:38 +0200, Andreas Unterkircher wrote: >> Hi Andreas, >> >> did the hints you received solve your problem? > > I don't have any globus-gma zombies on the CE. The load is simply high. > It seems to be related to Atlas' job distribution. Maybe they also > changed something. The job throughput is fairly high now. Even today > there are over 3500 finished Atlas jobs here. Yesterday there were 10000 > finished jobs. Probably this is somehow related. Did you receive my previous mail? Your CE is still flooded with grid_manager_monitor_agent processes. This is caused by bug #48588. You have to restart globus-gatekeeper. -- Cheers, Andrey Kiryanov.