CHPC Downtime: General environment clusters and Beehive and Narwhal Windows Servers - Tuesday, July 11, 2023 starting at 7:30am
Date Posted: June 26, 2023
General environment clusters – notchpeak, kingspeak, lonepeak, and ash, as well as the frisco nodes and the cryosparc nodes
The work on the clusters includes upgrading the OS to RockyLinux 8.8, along with updates to the lustre drivers, CUDA and ROCM. Users should log out of interactive nodes and close FastX sessions before the start of the downtime. A reservation is in place to drain the slurm batch queue of these clusters of running jobs by 7:30 am on July 11,2023. Any job that will not be able to complete before the start of the downtime will wait in the queue and be started after the downtime has been completed. Note that for the stand-alone cryosparc servers, the OS version will not be upgraded, but updates/patches will be applied, and all jobs running on these servers should be stopped before the start of the downtime. Expect the clusters to be unavailable for most of the day.
General environment ondemand servers
These will be updated to the latest version of ondemand (3.0).
The general environment /scratch/general/nfs1 file system and the cottonwood{9-17} file systems
These file systems will be updated one at a time. When a given file system is being updated, the file system will not be accessible. If you cannot access a given space on one of these file systems, please wait about 30 minutes and try again.
Beehive and Narwhal
Both systems will have OS updates applied. In addition, the version of ArcGIS will be updated from version 2 (2.5.2 on Narwhal and 2.9 on Beehive) to version 3.1 on both systems. Expect the servers to be unavailable for most of the day.