Quick answer: To protect VMs with a Nutanix protection domain, first add the second cluster as a remote site in Prism Element (Data Protection > Remote Site), then create the protection domain (+ Protection Domain > Async DR), add the VMs, and give it a snapshot schedule with a retention policy for both the local and the remote cluster. The shortest repeat interval for asynchronous replication is 1 hour, so in a disaster you can lose up to one schedule interval of data. This guide covers the legacy protection domains in Prism Element, and the steps follow Nutanix documentation and lab guides.
What a protection domain is
A protection domain is a group of VMs (or volume groups) on a Nutanix cluster that you snapshot and replicate together. It holds the schedule, the retention rules and the remote site that receives the copies. A remote site is the other Nutanix cluster that stores the replicated snapshots.
| Term | Meaning |
|---|---|
| Protection domain | The set of VMs that share one schedule and one retention policy |
| Remote site | The target cluster, added with its Prism Element virtual IP address |
| Schedule | How often snapshots are taken and sent, for example every hour |
| Consistency group | A subset of VMs in the domain that is snapshotted together, crash-consistently, as one unit |
| Async DR | Asynchronous replication, with a repeat interval of at least 1 hour |
Before you start
- Two Nutanix clusters that can reach each other over the network. Check the current Nutanix Ports and Protocols reference for the ports your AOS version needs between the sites, and open them on the firewalls in between.
- Prism Element access to both clusters with an account that can manage data protection.
- The remote cluster's virtual IP address (the cluster VIP that you use to browse to Prism Element).
- Enough space on the remote cluster for the replicated snapshots and your retention count.
- A license that covers your schedule. Asynchronous replication every hour or longer is available at the Starter level. Intervals of 1 to 15 minutes (NearSync) need a higher license tier.
Plan the network between the two sites before you replicate. The IPv4 subnet calculator helps you check that the management subnets and routes line up.
Step 1: Create the remote site
- In Prism Element on the first cluster, open Data Protection from the main menu.
- Select the Remote Site tab, then + Remote Site and choose Physical Cluster.
- Enter a name for the remote site and the virtual IP address of the remote cluster.
- Set the Network Mapping so VMs land on the right network after a failover, for example the source network (
AHV: base) mapped to the matching destination network. - Choose the source and destination containers (the vStore mapping; the default is
default-containeron both sides) and select Save. - Repeat the whole procedure on the other cluster, pointing back at the first one. Both clusters need a remote site entry for replication to work in both directions.
Step 2: Create the protection domain
- In Data Protection, select + Protection Domain and choose Async DR.
- Enter a name for the domain and select Create.
- In the list of unprotected VMs, select the VMs to protect, then choose Protect Selected Entities. To snapshot a subset of VMs together, put them in one consistency group at this point.
- Select Next to move to the schedule.
Step 3: Add a schedule and retention
- Select New Schedule.
- Choose how often to repeat. For asynchronous replication the minimum is every 1 hour. This interval is also your worst-case data loss if the production site fails.
- Under the retention policy, set Keep the last N snapshots separately for the local cluster and for the remote site. Remember the retention count multiplied by the interval gives you how far back you can restore.
- Tick the remote site so the snapshots are replicated, then select Create Schedule and Close.
A protection domain can have more than one schedule, for example an hourly schedule kept for a day and a daily schedule kept for a month.
Step 4: Check that replication works
Open the protection domain in Data Protection and watch the Replications and Snapshots tabs. After the first scheduled run, the snapshot should appear on both clusters, and the remote site should show the replicated copy. The first replication copies the full data of the protected VMs, so it takes longer than later runs, which only send changes.
Failover: Migrate or Activate
Use these options on the protection domain when you need the VMs on the other cluster:
| Action | When to use it | What happens |
|---|---|---|
| Migrate | Planned move, with both sites running | The domain moves to the other site and the VMs arrive powered off. Replication is reversed. |
| Activate | Disaster, when the production site is down | You run it on the surviving site. The VMs come up from the state of the last completed replication. |
After either action, power on the VMs on the destination cluster and check their networks. Test a planned migration during a maintenance window so you know the procedure and timing before you need it.
Frequently asked questions
How much data can I lose with asynchronous replication?
Up to the schedule interval. With an hourly schedule, VMs restored on the remote site can be up to an hour behind, plus the time a replication takes to finish.
What is the difference between a protection domain and a protection policy?
Protection domains are the Prism Element feature that this guide covers. Newer deployments managed from Prism Central use protection policies and recovery plans instead. Use whichever your environment and Nutanix documentation for your AOS version recommend.
Can one protection domain have several schedules?
Yes. You can add several schedules, for example a frequent one with short retention and a daily one kept for longer.
What is a consistency group?
A consistency group is a subset of the VMs in a protection domain that Nutanix snapshots together, crash-consistently, as one unit. Use it for VMs that depend on each other, such as an application server and its database.
Why does the remote site need to be configured on both clusters?
Replication and failback work in both directions, so each cluster needs to know the other. Configure the remote site on both sides before you create the protection domain.
Next steps
After the first successful replication, schedule a planned Migrate test and document your failover steps. For related reading, see the Data Center guides, the Oracle Linux Virtualization Manager install guide for another virtualization platform, and server room and data center design best practices.
Thank you for your message.