Nutanix: failover recovery with Async DR

by lionardo

There two scenarios for failover plan and migrating to DR:
First, when main site is online and have time to shutdown the servers properly.
Second is when main site offline for some reason and you can’t access to servers.

1: To DR and back online in 60 seconds

I have two sites: main and DR.
On main site connect to servers and shut them down (his step is not necessary, but recommended). When all machines are off, connect to PRISM and go to Data Protection – Async DR

Mark one server that you want to migrate to DR site and click on “Migrate”. If server was online and didn’t shut down before, Nutanix will automatically power him off

Connect to PRISM on DR site.
Wait until Nutanix copy server to DR site, you can track it in “Recent Tasks”, look for “VM destroy” completed to 100% in main site and “VM register” on DR site.

By default Nutanix keep servers in power off state after migration in case you need to change network or memory settings.
Final step is going to VM list and power on servers

To return servers to main site, migrate servers from DR to main. In this scenario, all servers will up to date to moment when we shut them down

2. When main site is offline

Connect to PRISM on DR site and go to Menu – Data protection, mark relevant data protection collection and chose Activate

Wait until task is complete, go to VM and power on servers that was in activaited data protection. The server will power on from last time that snapshot synced to DR

2.1 Back to production

Power on HOSTs on main site and wait until they online, the server will remain offline.
Connect with SSH to one of the HOSTs on main site and run this hidden CLI command: 

ncli pddeactivate_and_destroy_vms name=Data Protection Name

Replace “Data protection name” with relevant name to you. Type Y on warning, this will delete VM on main site.
After that go to Data protection tab on DR site, select data protection collection and hit Migrate.

This action will take some time and the VM will unreachable when migration is running. You can track migration status in Metrics tab on main and DR sites. At the end power on VM back online