Recovery Plans

A recovery plan is an ordered runbook that recovers a group of workloads together, in a priority order you define. Plans are the way to recover an entire application or site in one coordinated action, rather than one VM at a time.

Recovery plans are a DRM capability — they live in the DRM Recovery Plans view.

Why use a plan

• Recover many workloads with a single action.
• Control startup order so dependencies (databases, then application servers, then front-ends) come up in the right sequence.
• Rehearse the whole group with one test.
• Have the plan replicated to the recovery site so the surviving site can execute it even if the primary is gone.

Author a plan

1. Open the DRM Recovery Plans view and create a new plan.

2. Give it a name.

3. Add steps — one per workload — and for each step set:

   Step setting	Meaning
   Workload	The protected VM this step recovers
   Action	Failover (real recovery) or Test failover (isolated drill)
   Priority group	Lower groups run first; workloads in the same group run together
   Boot delay	Seconds to wait before starting this step, to stagger startup

4. Save the plan.

Order by dependency

Put foundational services (databases, directory services) in the lowest priority group so they start first, and add boot delays to give them time to become ready before dependent tiers start.

Plan replication and single-writer

When the sites are DRM-paired, a plan you author on the primary is replicated to the recovery site as a read-only copy. This is what lets the recovery site run the plan during a real disaster.

• On the recovery site, the replicated plan is read-only — it cannot be edited or deleted there. This "single-writer" rule prevents conflicting edits.
• Edit the plan only at the site that owns it; the changes replicate to the peer.

Run and test a plan

From the Recovery Plans view, each plan offers:

Control	What it does
Run plan	Executes the plan for real — fails over its workloads in priority order
Test	Runs an isolated, non-disruptive drill of the plan
Cleanup test	Tears down the isolated test environment
Run history	Shows previous runs and their outcomes

"Run plan" moves production

Run plan performs real failovers and moves the workloads to the recovery site. Use it only in an actual recovery, or on disposable workloads. To rehearse safely, use Test instead. See Failover & Recovery for the difference.

Disaster execution from the surviving site

If the primary site is lost, operate the recovery site's DRM. Because the plan was replicated there as a read-only copy, you can run the replicated plan from the surviving DRM to recover the group — without any dependency on the primary. The run is recorded in the recovery site's run history.