2

I have an existing SQL Server instance with three databases and a small amount of data.

I've been asked to create an Always On Availability Group now. The existing server would need to have the drives re-sized and only has 3 vCPUs.

So if you had a choice, what would be easiest/safest? Spin up the new servers, setup Always On, then migrate the data from the existing server?

Or should I keep the existing server and instance and add Always On Availability Groups to the existing SQL Server instance?

4

To migrate or not to migrate, that is the question: Whether 'tis nobler in the mind to suffer the slings and arrows of past setup sins and mistakes, or to take arms against that sea of troubles and by migrating end them.

Or not. If you were asking "Should I upgrade in place or migrate?" I'd have directed you to an older (and longer) answer I gave which, in short, says MIGRATE - ALWAYS MIGRATE!

But you didn't ask about an upgrade. You asked about an Availability Group. I commend you to that post on in-place vs migration upgrade decisions and contemplate where you are. If you are on a virtual and like the version of SQL you are on and only have to reconfigure "stuff" - I see no pressing reason to do a migration. You simply can add windows clustering, setup the AG feature, and start building your AG. You can use this as an excuse to cover for past sins because most config changes, etc. are "online" settings you can make as you go. On a VM you can play with your CPU config and RAM, etc.

If you need a new version of SQL Server or made some crucial installation mistake (hard to think of any, perhaps 64kB drive allocation size, but that's not the end of the world and you need new drives anyway), then sure install afresh and migrate. Or if you are on a physical, have the licensing budget, and can demonstrate the need for more cores, then you have no choice but to build new and migrate.

BUT! You are building an AG. You now have a magical THIRD OPTION. You could build your new server just the way you want it. You could add that as a node into your AG. You could build the AG with this instance participating as a replica in the AG with two instances you like the config and sizing of better. And then just failover to one of those and evict your current node and kick it goodbye. So you could use your AG to sail on over to the new hardware with minimal user frustration and look like a hero.

  • 1
    You get my vote for the size of the smile brought by your opening paragraph :D – youcantryreachingme Oct 8 at 23:27
  • Awe shucks. Thanks! :) – Mike Walsh Oct 8 at 23:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.