The advice below has been sent to the contact person for each VM and storage volume on the "qld" availability zone in the NeCTAR Research Cloud.
Dear NeCTAR User,
This is in relation to the email(s) you should recently have received about migrating your NeCTAR virtual machines (Instances) out of the DC2 data centre at St Lucia, Queensland. If you have not received this email, please contact <email@example.com> as soon as possible.
A recent NeCTAR database snapshot indicates that you (may) have one or more Volumes in the "qld" availability zone. Like the Instances, your Volumes also need to be moved out of DC2. The deadline of January 30th 2015 applies here as well.
If you no longer require the data in the Volumes, please just Delete them using the NeCTAR dashboard. This will help us when the time comes to do the final DC2 shutdown.
If you need to keep the data, then you will need to manually copy it to another location.
- Unfortunately, it is not possible to use the volume snapshot feature to perform this migration (due to size constraints).
- One option is to create a new Volume in some other availability zone (e.g. "QRIScloud"), attach the Volume to an Instance in that zone, and use a tool like "scp" or "rsync" to copy the data out of the "qld" Volume and into the new one. If you need extra temporary Volume storage quota in order to create the new Volume, please request it through the NeCTAR dashboard. Please also let us know via a support ticket so that we can assist in getting the quota change actioned.
- Other options include copying the data to an RDSI collection.
There are a couple of important issues that you need to take into account.
- There are known issues when Volumes have been attached to an Instance. Sometimes operator intervention is required. This can introduce significant delays. Please allow for this. If you experience problems with volumes, please raise a support ticket as described below.
- There is a known OpenStack bug that can cause issues if you attempt to Terminate an Instance that currently has a Volume attached to it. Always detach the Volume from the Instance and wait for a few minutes before you attempt to Terminate the Instance or delete the Volume. If the detach doesn't complete, please raise a support ticket.
- It is not currently possible to launch new Instances in DC2. This means that if you don't currently have any live VMs in DC2, you will not be able to copy out the data. If you are in this situation, please raise a support ticket, and we will try to work out a solution.
- When you are copying your data from DC2 to Polaris (or any other availability zone), your transfer will be competing with other transfers (e.g. of RDSI collections), resulting in the transfers running more slowly than usual. Please allow for this.
We therefore recommend that you start migrating your Volume(s) and associated VMs as soon as possible.
You can raise a QRISCloud support ticket by emailing <firstname.lastname@example.org> or alternatively by visiting the QRIScloud Support page at https://qriscloud.zendesk.com. If you have any queries related to this email, please raise them in a support ticket.