To Research Cloud users at the University of Melbourne,
Yesterday, Thursday June 01, between 10am and 12:30pm, the Ceph storage cluster in the Queensberry St Datacentre had a few service disruptions causing client IO requests to be blocked for several 10-minute long periods. This caused client IOs to the VMs and Cinder volume storage in the melbourne-qh2 and melbourne-qh2-uom AZs to be paused. Eventually the service disruptions were resolved and the paused client IOs came through. No data was lost on the cluster, and while some tenants would have experienced temporarily hung VMs or Cinder volumes, they should recover by themselves after the disruptions were resolved and the delayed requests got processed.
We sincerely apologise for any inconvenience caused. If you still experience any issue with your instance or storage, please log a ticket at https://support.ehelp.edu.au and we will assist.