Protecting Workloads

Protecting a workload means enrolling it for replication from the primary site to the recovery site. Once protected, its disks are copied to the recovery site on a schedule, producing recovery points you can fail over to.

You can protect a workload from either DRM (Workloads view) or Cockpit (a VM's Replication tab). Both create the same replication job.

Protect a workload from DRM

1. Open the DRM Workloads view. Unprotected VMs show an Unprotected badge and a Protect button.

2. Select Protect (or Enroll Workload) for the VM.

3. In the enrollment dialog, choose:

   Setting	Meaning
   Target site	The recovery site to replicate to
   Target datastore	The storage pool at the recovery site that will hold the replica
   RPO (minutes)	How often to replicate (how much data loss you can tolerate)
   Recovery points to keep	How many historical recovery points to retain at the DR site
   Compression	Compress replication traffic to save bandwidth

4. Confirm. DRM configures replication through Cockpit and starts the baseline (first, full) copy.

Protect a workload from Cockpit

1. Open the VM and go to its Replication tab.
2. Choose Configure and complete the same settings (target site, target datastore, RPO, recovery points, compression).
3. Save. The baseline sync begins.

The baseline sync

The first replication is a full copy of the workload's disks and can take time proportional to the disk size and link speed. Subsequent syncs are incremental — only changed blocks are sent.

When the baseline completes, you will see:

• At least one recovery point listed for the workload at the recovery site.
• A Last Successful Sync timestamp.
• A healthy replication status.

A recovery point means the replica is really there

A recovery point is only reported once the replica's disk files have physically landed and been verified at the recovery site. If a sync fails to transfer, it is reported honestly as a failure and does not create a usable recovery point.

Multi-disk workloads

VMs with more than one disk are fully supported. All of the workload's disks are replicated together and registered as a single recovery point — a failover boots the workload with every disk attached. A recovery point for a multi-disk VM is marked complete only when all of its disks have landed at the recovery site.

Choosing an RPO

The RPO is a trade-off between data protection and resource use:

• A shorter RPO (e.g. 5–15 minutes) means less potential data loss, but more frequent syncs and more network/storage activity.
• A longer RPO (e.g. 60 minutes or more) is lighter on resources but risks more data loss in a disaster.

You can change the RPO at any time — see Managing Replication.

Verify before you rely on it

After the baseline completes, run a test failover to confirm the workload actually boots at the recovery site. This is non-disruptive and is the only way to be sure the protection is real.