The abundance of options causes anxiety. That’s a proven fact. It is only true, however, if those options are hardly distinguishable. This is not the case with Nutanix DR options at all as each feature has a clear purpose and application.
It all starts with an SLA. Understanding your disaster scenarios and the amount of data that you are ready to lose in the event of the disaster is the key to choosing a solution that is right for you.
Roughly, there are several main options:
- Local backups (also known as Time Stream). Use these to roll back any Guest OS maintenance that went wrong. Remove a snapshot when the maintenance is over and the rollback is no longer required. Do not leave your snapshots unattended.
- Remote backup and DR:
- Asynchronous Replication: Protection from VM Corruption and Deletion as well as from a total site failure. Snapshots can be 1hr or longer apart.
- Near-sync Replication - leverages Lightweight Snapshots that are snapshots of the metadata. Recovery points can be 1-15 minutes apart.
- Metro Availability - Considered to be synchronous replication. So far only available on Hyper-V and ESXi.
Once you decide on the method, you can add complexity with architecture. Whether to keep a full-scale cluster to be able to withstand full site loss, or have a single node cluster at a remote site for off-site backups. Or maybe it is more convenient for you to leverage Cloud Connect into AWS or Azure Cloud, or sign up for Nutanix DRaaS using Xi Leap.
The choice is not that overwhelming after all.
Some reading:
Prism Element Data Protection Guide: Asynchronous Disaster Recovery between On-Premises Datacenters
KB-5159 NCC Health Check: hyperv_metro_setup_check
Nutanix University Tech TopX: Metro Availability