The Center for High Performance Computing will begin emergency maintenance to apply security updates to high-performance computing clusters (granite, notchpeak, kingspeak, lonepeak, and redwood) in a rolling fashion today (July 22). The updates are necessary to address a recently announced vulnerability. System administrators will update and reboot compute nodes when no jobs are running on them. This update will not affect running jobs, but researchers may see nodes in a “DRAIN” state before updates are applied; new jobs will not be able to start on a node until updates on that node finish.
The CHPC will also update interactive nodes, including login nodes (such as granite1 and granite2, redwood1 and redwood2, and similar nodes on other clusters) and research group-owned interactive nodes, on all high-performance computing clusters. Administrators will reboot the interactive nodes today at approximately 12:00 p.m. MDT to apply important security updates. We recommend signing out from interactive nodes prior to this time. Any active sessions or unsaved progress on interactive nodes will be lost when the nodes reboot.
In accordance with the CHPC’s emergency maintenance policy, important security patches must be applied promptly to protect system integrity and safety across our shared infrastructure. Thank you for your understanding and patience.
If you have any questions or concerns about upcoming security updates, please contact the CHPC by email at helpdesk@chpc.utah.edu.
