3.3 Planning for Failure

After workloads are properly protected, it is necessary to periodically test the protection and be ready for catastrophic system failure. The following tasks show how to be ready and respond.

3.3.1 Workload Testing

Use Test Failover to confirm the integrity of a replicated workload. Typical Test Failovers take only an hour.

  1. Click Workloads on the toolbar to navigate to the Workloads page.

  2. Select the check box next to the workload you want to test.

    Workload Commands are activated.

  3. Click Test Failover.

    The Command Confirmation page is displayed.

  4. Select a Recovery Point to use for the test, if desired.

  5. To run the Test Failover immediately, click Execute. To continue, see Step 6. To cancel the Test Failover, click Cancel.

    or

    To view and optionally change any workload settings before running the Test Failover, click the workload name. Click Test Failover at the bottom of the Web page to restart the Test Failover, then click Execute.

  6. Navigate to the Workloads page. When the workload’s Replication Statuschanges to Live, the virtual machine is running.

  7. Log in to the virtual machine by using Remote Desktop Connection or the VMware Virtual Infrastructure Client.

  8. Inspect the virtual machine to verify that the recovery workload is functioning properly.

  9. Shut down the virtual machine.

  10. Navigate to the Dashboard. Select Mark Test as Success or Mark Test as Failure for the workload.

NOTE:PlateSpin Forge logs the success or failure of the test failover as an event. This event can later be retrieved as part of an Events report.

3.3.2 Preparing for Failover

When a protected workload fails, Preparing for Failover protects your workload, but gives you time to determine if the problem is trivial.

  1. Click Workloads on the toolbar to navigate to the Workloads page.

  2. Select the workload that requires a Failover.

  3. Select Prepare for Failover.

    The Command Confirmation page is displayed.

  4. Select a Recovery Point to use for the test, if desired.

  5. To run the Prepare for Failover immediately, click Execute. To continue, see Step 6. To cancel the Failover, click Cancel.

    or

    To view and optionally change any workload settings before running Prepare for Failover, click the workload name. Click Test Failover at the bottom of the Web page to restart the Test Failover, then click Execute.

  6. Navigate to the Workloads page. The workload Replication Status changes to Preparing Failover.

  7. When the recovery workload is running and ready to be joined to the main network, its Replication Status changes to Failover Prepared.

3.3.3 Running a Prepared Failover

After preparing for a failover, carefully investigate the situation to make sure a failover is required. If it is required, bring the recovered workload on to the main network by running Failover.

  1. Click Workloads on the toolbar to navigate to the Workloads page.

  2. Select the workload that requires a failover.

  3. Select Failover.

    The Command Confirmation page is displayed.

  4. Select a Recovery Point to use for the test, if desired.

  5. The workload Replication Status changes to Going Live.

    When the operation is complete and the recovery workload has taken the primary workload’s place on the network, the Replication Status changes to Live.

3.3.4 Running an Unprepared Failover

In the rare case, such as natural disaster, where it is immediately obvious that a primary workload is irreparably disabled, run Failover directly.

  1. Click Workloads on the toolbar to navigate to the Workloads page.

  2. Select the workload that requires a failover.

  3. Select Failover.

  4. To run the Failover immediately, click Execute. To continue, see Step 5. To cancel the Failover, click Cancel.

    or

    To view and optionally change any workload settings before running the Failover, click the workload name. Click Failover at the bottom of the Web page to restart the Failover, then click Execute.

  5. Navigate to the Workloads page. The workload Replication Status changes to Running Failover.

    When the failover is complete, the Replication Status changes to Live. The workload is now running on the PlateSpin Forge Appliance Host.

3.3.5 Removing a Snapshot After Failover (Optional)

Forge leverages hypervisor snapshot technology to facilitate a rapid failover. Although it is not necessary, snapshots can be removed after the failover to improve performance and reclaim disk space.

NOTE:While the snapshot is being removed, the responsiveness of the workload is temporarily degraded. It is recommended that you perform this process during a scheduled down time.

  1. Login to the VMware Infrastructure Client using the Host Appliance IP address. Ignore any warnings.

  2. Expand the Virtual Machine list if necessary and locate the VM with the snapshot you want to remove.

    NOTE:While it is not necessary, the removal is significantly faster when the host VM is powered off.

  3. Right-click the VM and select Snapshot > Snapshot Manager.

  4. Select the snapshot you would like to remove and click Delete All.

    All changes are merged into the base.