Could you extend the capacity extender to the servers?
There are at least two use cases.
1: If there is a hardware problem at the data center and a rack, or set or rack, or even a zone goes down.
2: If you need to reboot a server for maintenance or any kind of OS issues.
Lately Azure has not let us purchase capacity reservations on specific skus so we cannot guarantee that our VM's will come up after a reboot.
It would be nice for Azure to automatically try other skus for us just to get the server back online.
However, being that these are servers, and we most likely have RI's in place, there should be a maintenance window we can set as we are setting up this capacity extender, that will allow Nerdio to attempt to change the sku back to the default sku as to prevent the possibility of double billing.
Capacity Extender for Servers
Ooo great idea!
Maybe if we combine this idea and this idea (Azure Capacity Extender - Return to Default SKU & clear order sequence of booting alternate SKU – Nerdio Help Center) from Ken where it will try to go back to the default SKU on a timeline, it will do what you're needing.
You suggestion for a Maintenance Window would be a good addition.
If you're doing the resize, there's a potential of the VM not being able to boot.
Do you think you'd need/want to create a temporary VM just to see if the SKU is available or does the existing Nerdio Azure Capacity Extender logic work for what you're needing?
Yes, I have done this in the past. Last October ('24) when we had several clients in the South Central region I would quickly attempt to spink up a VM on a sku I wanted to resize a VM to as a test. If it failed to build, I wouldn't even try to resize the VM.
If this could be automated, that would make the process so much easier!
Sweet!
We should be able to copy the temp VM logic from another part of NMM to make this easier to implement.
Please sign in to leave a comment.
Comments (3 comments)