Create a new patched instance and auto swap instead of bringing down single instance production deployments
When running a single instance every so often the instance will go down for patching. The instance appears as "VM Stopped" for a period of time, then comes back.
Rather than forcing downtime to a production role, why not spin up a patched version of the instance then swap it in for the unpatched instance, and delete the old one?
MS documentation already points out the instances can be created and destroyed at any time, so we've already created our apps in a way that can handle such scenarios. Applying the patches this way will save the downtime (we've observed up to 30-40 minutes in some cases).
One complexity here is because these instances are persistent, creating a snapped copy might be tricky or impossible.
Is this something that could be enabled with Auto-Scale?