This is a reminder that our next HBSGrid maintenance is scheduled for Monday, February 1st, 8 am - 12 pm. This is mostly a behind-the-scenes maintenance with no major changes of concern for all our users, and will likely finish well in advance of the 12p window end.
How will this affect you?
- You will not be able to log in to the HBSGrid via terminal or NoMachine
- All NoMachine sessions will be terminated (both active & disconnected sessions)
- Interactive jobs will be terminated
- All other running jobs may be temporarily suspended. There is a small chance that your jobs may fail when restarted, though this is unlikely.
- Jobs without explicit software loads will be handled more gracefully
bjobscommands using the
-l(long detail) option will run more quickly
- The system will no longer leave core dumps in directories when programs crash
Files in the scratch space (
/export/scratch/) will be deleted if they haven’t been modified in over 60 days
What work is being performed?
- NoMachine will be reconfigured & restarted to prevent sharing local disks publicly across the login nodes
/usr/local/binon compute nodes will point to a central storage mount to handle jobs where module loads or
PATHadjustments were omitted
LSF's job log rotation parameter
MAX_JOB_NUMwill be updated
ulimitoption will be adjusted to prevent writing core dumps
- We will run the scratch filesystem retention script, which deletes files and empty directories that haven't been modified in the last 60 days
We thank you for your anticipated cooperation as we make improvements to our research computing environment. Our next maintenance will occur on Monday, March 1st. If you have any questions/comments, please email us at email@example.com.