If your jobs are stuck in the queue forever (e.g. other users hogging the CPUs, jumping the queue, etc.) then you should do the following:
- Talk to the offending users directly. Often, they work in the office next to you...
Use the Mailing_list to send email to all cluster users and to discuss the problem.
As a last resort, see Joshua.
If you have general suggestions or specific but general problems with the queuing system, please add your information to: queue-structure.
- Please change the disk usage page, so that:
- panfs and netapp are separate
- The usage is sorted by size
If chef is inaccessible or you encounter issues with job submissions, then contact Joshua.