First failover

Datamotive supports two types of failover: a DR drill (non-disruptive, isolated network) and a production failover (brings workloads live at the target). Start with a DR drill to validate your plan before you need it in a real incident.

DR drill (recommended first test)

A DR drill boots the shadow VMs in an isolated test network. The source workload continues replicating normally. No data is modified at the source.

1. Open the plan and click DR Drill

   Navigate to Plans, select the plan you want to test, and click DR Drill on the plan detail page.

2. Select the recovery point

   Choose the recovery point to boot from. The default is the latest available point. You can select an older point to test a specific point-in-time state.

3. Confirm the test network

   Datamotive creates an isolated VPC or VLAN for the drill. The shadow VMs are connected to this network and cannot reach production systems. Confirm the network configuration and click Start drill.

4. Wait for VMs to boot

   Shadow VMs boot from the recovery point. Boot time is typically under 3 minutes per VM. Monitor progress on the Jobs page.

5. Validate the application

   Connect to the shadow VMs using the test network addresses shown in the drill detail panel. Validate that:

   • The OS boots without errors
   • Application services start (web server, database, etc.)
   • Data is consistent with the recovery point timestamp
6. End the drill

   After validation, click End drill. Datamotive powers down the shadow VMs and removes the test network. Replication continues without interruption. A drill report is generated automatically.

Production failover

A production failover is used during an actual disaster or a planned maintenance migration. It brings shadow VMs live at the target site on the production network.

1. Confirm the incident

   Before triggering a production failover, confirm that the source site is actually unavailable or that a planned maintenance window is in progress.

2. Open the plan and click Failover

   Navigate to Plans, select the plan, and click Failover. Select Production failover when prompted.

3. Select the recovery point

   Choose the most recent recovery point, or select an earlier clean point if the latest is suspected to be corrupt (for example, in a ransomware scenario).

4. Review and confirm network mappings

   Datamotive shows the source-to-target network mapping. Verify IP assignments and DNS settings. Make any needed adjustments before confirming.

5. Initiate failover

   Click Confirm and fail over. Datamotive boots the shadow VMs on the production target network and applies network customizations (hostname, IP, DNS). Boot and customization take approximately 5 to 10 minutes per VM.

6. Validate and update DNS

   Test the recovered workloads at the target. Update DNS records to point to the target IP addresses. Notify application owners that workloads are online at the DR site.

After a production failover

Once workloads are running at the target, you have two options:

• Fail back: When the source site is repaired, reverse replication to synchronize changes made at the DR site back to the source, then cut over again. See Failback.
• Declare the target as the new primary: Reconfigure the plan so the original target becomes the source and the repaired site becomes the new DR target.

Related docs

• Failover workflow

• Failback workflow

• DR drill workflow

• DR drills best practices