Reducing cost during NFA to NMM Migration

Hello Nerdio partners!

As you all know NFA will be riding off into the sunset early next year. There’s a guide for migrating that you can find here, but one concern I’ve heard from some of you is that the migration process can incur extra costs due to having simultaneously running environments. There’s definitely some merit to that concern, so I’d like to take a minute to provide some tips that can help reduce that extra cost.

1. Keep your session hosts in the new environment powered down unless you are actively testing.
   
   This is the biggest offender. Having two AVD pools running simultaneously can double costs in some cases. The easiest way to handle this initially is to just disable auto-scale and provision a single host manually. That way you can power it on and off as needed to test and you’ll only be billed for the time it’s running.
   
   If you need to test auto-scale, you can enable it, configure it as needed, then shut it off when you’re done testing. Your settings will be preserved for when you need to turn it on next.
   
   
2. Only assign users to desktops if you will be actively testing with them.
   
   Nerdio bills on a per user basis. If you have users assigned to desktops in NMM and in NFA, you’ll be billed on both platforms. There are a couple of easy ways to prevent being double billed:
   1. Create a new user that’s specifically for testing the new environment
   2. Once you have a pool created, have a pilot user or set of pilot users that have their NFA desktops removed and are only assigned to a desktop in NMM
3. Read through the documentation ahead of time and only spin up resources once you’ve fully planned out how it will go. Having a game plan will reduce the amount of testing you’ll need to do and minimize any extra compute cost.

Let us know if these tips help you during your migration, and if you have any additional advice for partners that you learned while migrating, feel free to comment below!