Cloud-Based elements to keep track of include the many new aspects of managing your servers and instances once you move to the cloud:
– What is your maintenance window, does it impact your applications, and do your users know about it? Make sure they know the window is there, how it will impact their work and what it looks like when things are in maintenance mode. It’s easy to overlook notifying and working with users on the systems – be sure they’re informed. As part of your process, you’ll want to make sure you have an "all clear" notification as part of the process.
– What is your recovery process if you have to restore systems? How is it different from your existing on-premises routines? Make sure those different steps are documented for your recovery team and that they understand the differences between your systems. Test. Do a restore/recovery and make sure you know how it works, the time required and the permissions and steps required.
– What is your firewall check process? Specifically how often do you audit the people and machines that are allowed through your firewall at the provider? Have you checked to make sure all of those access points are necessary? This is true as well for the SQL Server instances – who has access, why, and is it still needed? This can have an added layer of complexity in the cloud because you may have a different team watching over your systems than you may have on the on-premise systems.
– What performance checks do you have in place, and what are the *expected* values for those metrics? These may be vastly different in the cloud – from the tools you’ll use to the expected values. It’s important to make sure all of your team knows the specifics.
– If you have systems that point to and use your SQL Server instance(s), what happens if you have to rebuild an instance? In particular, are DNS updates needed, do load balancers have to be changed, are there permissions to be updated? It’s also possible that applications will need attention, from connection strings to other values and setup options.
– Permissions and access – it’s quite possible that access to your provider’s tools is done to some extent with web-based tools. Make sure your team has appropriate rights to access those tools and that everyone knows how to get into the systems and knows how they work. It’s easy to pay more attention to remote desktop access, while missing access through the tools that are part of the management console.
It’s important to know the tools, have proper access and experience with them and not be the only one with that experience. Sure, this is similar to the systems you have in-house or are working with on a colocation basis, but there are different tools, different access points and, in some cases different responses from the underlying systems that need to be added to your checklists.
What did I miss?