06/19/2023
Ceres cluster maintenance is scheduled for the week of June 19, to update system software. The cluster will be down for several days.
This will be a major upgrade. During the maintenance Ceres will be upgraded from Enterprise Linux (EL) 7 to EL 9. This will affect not only the system software but also the software available via modules. If you have installed software on your own, you may need to rebuild it under the new OS as well. Multiple emails have been sent to SCINet Users advising to test new software prior to the maintenance on the EL9 sandbox (login-dev.scinet.usda.gov).
We will also upgrade Globus.
The cluster will be released as soon as the upgrade and testing are done.
Queued jobs both on Ceres and the EL9 sandbox will not start if they cannot complete by 6:00 AM on June 19. In the output of the squeue command the reason for those jobs will be (ReqNodeNotAvail, Reserved for maintenance). These jobs will start after the scheduled outage completes.
If your jobs don’t need the default partition time limit, you can specify a shorter time limit using “–time=” option, see Ceres Job Script Generator. The Atlas cluster will stay up and running during the Ceres downtime. All Ceres users can run jobs on Atlas. If you have questions or need help, please email us at scinet_vrsc@usda.gov.