UPDATE: Regular scheduling on Oakley has been returned to service.
pdsh -j is broken on Oakley. It was broken by updates during the September downtime. We are currently working on resolving the issue.
Users who require a workaround should contact OSC Help for assistance.
Currently, the estimated RU charge reported at the end of a job shows an incorrect value for serial jobs on Oakley of the entire node. Jobs are being charged the correct amount in the official accounting records.
We have temporarily suspended scheduling due to some problems with the parallel scratch file system.
On April 17th 2013, at roughly 2PM, we will be rebooting the batch server on the Oakley cluster. Running jobs will not be affected, but there will be a brief disruption in scheduling, as well as the functionality of programs that communicate with the batch server (such as qsub or qstat) during the duration of the reboot.
This disruption is expected to be very short, and will be transparent to users unless you attempt to submit a job or view the queue during the outage. No other service disruptions are anticipated.