13:30 - 22nd Jan 2021
All impacted users have been emailed regarding the emergency maintenance of KVM6 and VZ25-SSD (Both nodes run on the same physical hardware/blade) upon further investigation, it looks like the issue locking the CPU threads to 800MHz may be due to not enough power being drawn from the PSU, there are some PSU failure notices that coincide with the time the issues started and the redundant PSU has taken over.
Advice is being sought from the datacenter on next steps, in the meantime, services have been restarted to minimise disruption however further reboots may be required.
Sorry for the inconvenience, further updates will follow when available
22:00 - 22nd Jan 2021
The issue was in fact caused by power distribution, with the careful ordering of redundant component swaps it was possible for a super micro engineer to replace the failed parts without impacting service further.
The issue is now resolved, any services that did not auto-start after the last reboot can be started from the VPS control panel.
Friday, January 22, 2021