I do not understand how to properly handle disaster recovery (ex: my Azure Datacenter unavailable) with Virtual Machines.
For example, SqlAzure makes it pretty simple: with the Geo-Restore feature, you can restore your database to your "secondary" datacenter and that's it.
We have a VM architecture to support Remote App( DC,Gateway,Broker, Session Host, etc), basically an infreastructure to support an application.
If the datacenter becomes unavailable, I know how to recover the data with SqlAzure, but I don't know the process to recover my infrastructure.
The best thing I can think of is: having your vhd's with geo-redundant store, your vhd's should be available in case of datacenter unavailable. Then you would add this vhd's as images using the portal, and then re create manually your infrastructure using this vhd's.
Is there a better way to do this?