The job scheduler currently allows general access for all users with a valid “hpc” account to standard computing resources available through the “batch” queue.
Each user may have up to fifty (50) jobs to the queue at any given time; up to twenty-five (25) of these queued jobs may be running. Job execution time (“walltime”) may be no longer than 240 hours (ten days). The maximum nodes accessible by any user can be capped based on your allocation and scheduler workload.
Jobs requiring access to resources outside of these constraints should request a reservation by contacting email@example.com.
Fairshare access based on departmental investment is currently being implemented.