forums

[Scheduled Maintenance] Decommissioning of old compute hosts in melbourne-qh2 zone 2019-06-03 to 2019-07-01

About 


Services affected: Nectar Research Cloud - melbourne-qh2 zone

Outage Window: Maintenance will be performed on compute hosts in a rolling fashion from 2019-06-03 09:00 AM AEST to 2019-07-01 18:00 PM AEDT 

Affected Availability Zones: melbourne-qh2


Reasons 


Research Platform Services (ResPlat) advise that we are decommissioning our end-of-life Supermicro cloud compute hosts in melbourne-qh2 availability zone. They are being replaced by newer compute hosts with considerable improvements in performance. Performance comparisons are available here: https://docs.cloud.unimelb.edu.au/benchmarks/spring_2019 (see “NeCTAR melbourne-qh2" vs “NeCTAR Supermicro”).


Detail 


The old equipment will be decommissioned and powered off on Monday 2019-07-01.


Between now and then, our Research Cloud support staff will be contacting users with instances on the end-of-life compute hosts to notify them when each of their instances will be migrated to the new compute hosts. Active instances will be shutdown, migrated to a new compute host, and then started again. The migrations will be performed on each of the end-of-life compute hosts in a rolling fashion and will take place between 3 June 2019 and 01 July 2019.


What Users Need To Do


To minimise risks to your instances, please ensure important data is backed up, and critical instances are snapshotted. If your instances are no longer in use, please delete/terminate them.  


Guide for snapshotting and backup of cloud instances: https://support.ehelp.edu.au/solution/articles/6000175790-snapshots  


Contact Details


If you require help regarding this scheduled maintenance, please contact us via rc-support@unimelb.edu.au or call (+61 3) 834 40999 and ask for the Research Cloud support.

Login to post a comment