MicrosoftI think this is really great, and I'm glad to see it finally come around. I was wondering when it would happen. :-) Just wanted to give you my thoughts on the managed instances.
How would you do DR testing in a farm that uses SQL Managed Instances?
I'm currently in a project where we're designing a large, consolidated SharePoint environment in Azure Gov with multiple business-critical large farms. I would be very interested in consider putting this in the design vs. building our own AAGs, but I'm concerned about how we'd test DR. I know that SQL Managed Instance is a managed service and thus DR is Microsoft's problem. However, they're being married to IaaS VMs which are owned, managed, and protected by the customer. How do we test the recovery of the SharePoint farm if the databases are in a Managed Instance?
I think it can be a great solution for DR itself. After all, you would only need to worry about protecting the application servers. However, how would you perform full DR tests? Unless I'm not seeing something, you'd have to do a live DR test in production since you only have the Managed Instances. I'd posit that this simply is not realistic. None of my customers would risk doing a live production DR failover just for the sake of testing DR; there must be no risk to production. So it appears a SQL Managed Instance could significantly hamper a customer's ability to perform full-environment SharePoint DR tests without live failovers. Perhaps there's a way to restore from one Managed Instance to another?
I'm sure I'm not seeing some option, so I'd be interested in solutions to the (perceived) DR problem. Yes, we should be able to trust our DR solution, but it takes another level of faith to actually perform it in production. Anyway, this topic is something to consider going forward as you build customer guidance.
HTH!
