DOWNTIME: Select CHPC systems, Tuesday, March 22, 2022 starting at 7am
Date Posted: March 8, 2022
CHPC will have a downtime on Tuesday,March 22, 2022 starting at 7 am.
This downtime will be used to
- update the OS on the windows servers Beehive and Narwhal
- update drive firmware on select storage trays
- make a network switch change in the ScienceDMZ to prepare for the new home and scratch storage
- start the process of the UPS update in the INSCC machine room
- start on the OS upgrade to RockyLinux8. During this downtime, the compute and interactive nodes of lonepeak and the frisco nodes will be updated
Impact:
- At about 7 am, for about 30 minutes, there will be no internet access (both wired and wireless) in the INSCC building.
- Starting at 8 am, Narwhal and Beehive will be down most of the day for OS updates.
- Starting at 8 am, the compute and interactive nodes of lonepeak and the frisco nodes will be moved to RockyLinux8. These resources will be unavailable for most of the day.
- A reservation is in place to drain lonepeak of running jobs before the start of the downtime.
- Pending jobs will remain in the queue and will run after the downtime is complete.
- Users should save any output and close sessions, including ssh, srun, ondemand and fastX sessions, on the these nodes before the start of the downtime.
- At about 9 am, the data transfer nodes (DTNs) on the Science DMZ (dtn{05-08}, sdss-dtn) will be unavailable for about 2 hours.
- /scratch/general/nfs1 in the general environment and /scratch/general/pe-nfs1 in the protected environment (PE) will be unavailable for a short window (less than 30 minutes expected). The outage will be short and jobs with I/O to these scratch file systems should be able to run through the outage.
- Group spaces on cottonwood-vg9 through vg14file systems will be unavailable for a short window. Again, the outage will be short and jobs with I/O to these spaces should be able to run through the outage.
- You can determine which file system your group space(s) are on by going to a cluster interactive node, cd to the group space in question, and doing a "df -h | grep groupspacename". this output will give you the file system such as cottonwood-vg9-1-lv1. Anything with vg9, vg10, vg11, vg12, vg13 or vg14 will be impacted.
Please let us know, via helpdesk@chpc.utah.edu, if you have any questions or concerns.