[Complete] Melbourne-qh2 availability zone Jan 05, 2016, 10:30am-11:30am AEDT
L
Linh Vu
started a topic
over 8 years ago
Update 01:05PM AEDT Tue Jan 5th 2016: the spanning tree issue has been resolved, Melbourne QH2 network is back to operational.
Update 12:50PM AEDT Tue Jan 5th 2016: the switches are running into some spanning tree issues, which are causing network disruptions in Melbourne QH2 Datacenter. The University of Melbourne Network Services team are working on fixing them.
Please be aware there will be a scheduled outage in the melbourne-qh2 availability zone of the NeCTAR Research Cloud:
Duration: 1 hour Start Time: 10:30 AEDT Tue Jan 5th 2016 End Time: 11:30 AEDT Tue Jan 5th 2016
The outage is required for the University of Melbourne Network Services to do firmware upgrade on our switches to enable support for new cloud infrastructure equipments and for the future separation of enterprise and research networks. This scheduled event is to finalise the previous upgrade on November 10, 2015, that Network Services did not fully complete due to an unexpected equipment failure.
Prior to the outage, if you have running instances in the melbourne-qh2 availability zone, you will receive a first notice 2 weeks out, then a reminder notice 1 week out, and a final reminder the day before, to let you know of the impact and actions you should take. While we expect no data loss to occur, we encourage you to take snapshots of your instances prior to the outage, for added protection.
At the start of the outage window, Melbourne Node will need to shut down all VMs in the melbourne-qh2 availability zone, then the compute nodes hosting them, to prevent data loss. Network Services will take over, flash new firmware on the switches and reboot them. Once testing has been completed, Melbourne Node will turn on the compute nodes, restart previously active VMs and notify users via email.
The other University of Melbourne availability zones melbourne-np and melbourne-qh2-uom are unaffected.
We apologise for the inconveniences and will aim to restore services as soon as possible.
Linh Vu
Update 01:05PM AEDT Tue Jan 5th 2016: the spanning tree issue has been resolved, Melbourne QH2 network is back to operational.
Update 12:50PM AEDT Tue Jan 5th 2016: the switches are running into some spanning tree issues, which are causing network disruptions in Melbourne QH2 Datacenter. The University of Melbourne Network Services team are working on fixing them.
Please be aware there will be a scheduled outage in the melbourne-qh2 availability zone of the NeCTAR Research Cloud:
Duration: 1 hour
Start Time: 10:30 AEDT Tue Jan 5th 2016
End Time: 11:30 AEDT Tue Jan 5th 2016
The outage is required for the University of Melbourne Network Services to do firmware upgrade on our switches to enable support for new cloud infrastructure equipments and for the future separation of enterprise and research networks. This scheduled event is to finalise the previous upgrade on November 10, 2015, that Network Services did not fully complete due to an unexpected equipment failure.
Prior to the outage, if you have running instances in the melbourne-qh2 availability zone, you will receive a first notice 2 weeks out, then a reminder notice 1 week out, and a final reminder the day before, to let you know of the impact and actions you should take. While we expect no data loss to occur, we encourage you to take snapshots of your instances prior to the outage, for added protection.
At the start of the outage window, Melbourne Node will need to shut down all VMs in the melbourne-qh2 availability zone, then the compute nodes hosting them, to prevent data loss. Network Services will take over, flash new firmware on the switches and reboot them. Once testing has been completed, Melbourne Node will turn on the compute nodes, restart previously active VMs and notify users via email.
The other University of Melbourne availability zones melbourne-np and melbourne-qh2-uom are unaffected.
We apologise for the inconveniences and will aim to restore services as soon as possible.