When a disaster occurs on primary location, you recover the Nova instances and Cinder volumes in the target disaster recovery site using SRM.

Note: You do not have to recover virtual machine of the Cinder volume in attached status.

Procedure

  1. In the vSphere Client of the target site, create a folder named OpenStack under the data center where the Compute Nodes are hosted.
  2. Under the OpenStack folder, create folders named like ’Project (<project-id>)’ for projects that include the recovered instances or volumes.
  3. Under each project folder, create an Instances folder and a Volumes folder.
  4. Log in to the Site Recovery manager in the target site.
  5. Under Replications, select the virtual machine and click RECOVER.
  6. Under Recover virtual machine, select Recovery options.
    1. Select Synchronize recent changes.
    2. Unselect Power on the virtual machine after recovery.
    3. Click Next.
  7. In the Folder section, choose Instances folder for virtual machine, and choose Volumes folder for Cinder volume shadow VM under it's associated project folder at the recovery site. Click Next.
  8. In the Resource section, specify the resource for the virtual machine at the recovery site and click Next.
    Note: The target cluster must have one-to-one mapping with the source compute cluster.
  9. In the Ready to complete section, review your selected settings and click Finish.
  10. For Nova instance VM, select the network from the Network adapter 1 tab and select Connect At Power On. For Cinder volume shadow VM, you can skip this step.
  11. Power on the Nova instance VMs. For Cinder volume shadow VM, leave it as power off.