University of Cambridge Computer Laboratory - Departmental GPU cluster storage maintenance – Maintenance details

All systems operational

Departmental GPU cluster storage maintenance

Completed
Scheduled for 18 December, 2025 at 16:30 – 17:38UTC

Affects

Virtual Machine Hosting

Under maintenance from 4:30 PM to 5:38 PM

GPUs

Under maintenance from 4:30 PM to 5:38 PM

Data Storage

Under maintenance from 4:30 PM to 5:38 PM

Other Secondary Storage Systems

Under maintenance from 4:30 PM to 5:38 PM

Updates
  • Completed
    18 December, 2025 at 17:38UTC
    Completed
    18 December, 2025 at 17:38UTC

    Maintenance has completed successfully.

    The shared development servers are available again.

    VMs can be started again via Xen Orchestra.

  • In progress
    18 December, 2025 at 16:30UTC
    In progress
    18 December, 2025 at 16:30UTC
    Maintenance is now in progress
  • Planned
    18 December, 2025 at 16:30UTC
    Planned
    18 December, 2025 at 16:30UTC

    We have been advised by the manufacturer that the storage server that holds home directories and VM disks on the departmental GPU cluster needs an urgent firmware update in order to avoid a possible data-loss issue.

    Due to the risk of data loss and the upcoming Christmas closure and staff leave, we think it's best to do this update at short notice on Thursday 18th December.

    This will affect personal GPU/CPU development VMs named dev-gpu-* and dev-cpu-*, as well as the shared development servers dev-cpu-1, dev-gpu-2, dev-gpu-acs and dev-cpu-acs.

    The shared servers will be shut down before the maintenance, and started again once it is complete.

    Affected VMs will be shut down before the maintenance, and will not be started again automatically. You will be able to start them again via Xen Orchestra once the maintenance has completed - please check this page and await confirmation that it's OK to try to start your VM.