Code.rc (the Research Computing GitLab service) will be unavailable from Friday, 2026-03-13 at 5:00 PM, through Sunday, 2026-03-15 at 11:59 PM, for planned maintenance. We expect to finish before the end of the window but have reserved the duration due to the nature of the changes.
We will be moving the Cheaha head node from GPFS4 to GPFS5. Maintenance is scheduled from 5:00 PM January 30th, 2026 to 6:00 AM January 31st, 2026. During this time Cheaha will be unavailable. This move marks the final part of our storage infrastructure upgrade and data migration.
How will this impact your work?
Please do not use technologies like VSCode "Remote - SSH" or Cursor Server to connect to Cheaha. These methods run all of their processes on the login node, which are automatically terminated.
Cheaha compute nodes located at the Technology Innovation Center on campus will be undergoing maintenance from June 16-22, resulting in reduced CPU and GPU capacity during those days, as well as reduced memory limits for jobs. Please plan Research Computing needs around the maintenance window.
We are asking folks who use our services to complete our Research Computing Survey. We want the best information possible on how users enjoy the services we offer and what ideas you have for the future. Please go to https://uab.co1.qualtrics.com/jfe/form/SV_dbaFBqVXIVkSbum to complete the survey by Friday, May 16 – it should take fewer than Five (5) minutes.
Cloud.rc (OpenStack) and Cheaha nodes within DC Blox will be undergoing maintenance during the week of April 21, 2025. Cloud.rc and Cheaha "amperenodes,” "pascalnodes,” and "intel-dcb" partitions will be inaccessible during maintenance. Note this means all GPU nodes will be inaccessible. Planned growth requires us to move hardware within the DC BLOX data center. Please plan your Research Computing needs around this maintenance window. If you need additional assistance, please reply to this email or reach out to support@listserv.uab.edu.
We are aware of increased transfer times into and out of LTS when using Globus. Transfer times are taking substantially longer than expected. The issue is occurring because of a bug in our version of Ceph, the LTS backend storage system. A bugfix is in the works by the vendor and we will communicate once we know more.
Note
This window applies only to researchers using GPFS5.
We will be performing upgrades on the GPFS5 storage system to optimize stability. During this process, researchers on the GPFS5 side will be unable to login via SSH, and Open OnDemand (https://rc.uab.edu) will appear unresponsive.
The maintenance window is scheduled for the morning of Tuesday, September 16, 2025, from 6:00 AM through 6:00 PM. We expect the process to take no more than 12 hours. We will restore service as soon as work is complete.