Virtual Machines managed by vRealize Automation are associated with a vRealize Automation reservation. Each reservation resides on a vRealize Automation compute resource that is associated with a vRealize Automation business group. These resource virtual machines can be protected by Site Recovery Manager.

Each virtual machine which is protected using Site Recovery Manager is assigned to a protection group and a recovery plan. When Site Recovery Manager fails over a protected virtual machine from the protected site to a recovery site, the virtual machines change clusters characteristics placed on the mapped datastore. After a failover occurs, a scheduled vRealize Automation data collection is execute against the recovery site, the virtual machine entry in the vRealize Automation database is updated with the new cluster, storage, and vSphere ManagedObjectReference. However, the reservation is not modified. This failure to modify the reservation might result in an error for any ongoing vRealize Automation resource actions, such as power state change, and the task might not be performed. To resolve this issue, each virtual machine must be updated within vRealize Automation to point to the new reservation.

Change the reservation for the virtual machines within vRealize Automation by using any of the following methods:
  • Select the Infrastructure tab and select Machines > Managed Machines > Change Reservation to update the reservation for each machine manually.
  • Automate the update of the reservation details dynamically:
    1. Configure vRealize Automation endpoints with a custom property
    2. Configure CloudClient scripts on the Site Recovery Manager protected and recovery servers
    3. Configure Site Recovery Manager to invoke CloudClient scripts during the execution of the recovery plan.
    Note: For CloudClient to automatically update the reservation after recovery, the recovery vCenter endpoint should be configured with reservation that includes the arrays used for the reservation.
  • For array-based replication, you should execute at least one sample recovery to be able to create the proper reservation for the recovery vCenter. The replication storage is not available on the recovery site, until you do so. Following are the additional steps required
    1. Execute array-based replication so that the storage is available on the recovery site.
    2. Perform a data collection for the recovery site in vRealize Automation.
      Note: After you restore on the recovery site, you must wait for all the data collection to complete before provisioning any VMs.
    3. Edit the reservation for the recovery site by adding the replication storage.
    4. Reprotect and execute a recovery back to the protected site and reprotect again.
    5. In vCenter, navigate to Site Recovery > Sites > Advanced Settings > Storage Provider and set property storageProvider.fixRecoveredDatastoreNames to true. This prevents the array-based replication to change the name of the storage when a recovery is executed.
Note: This document describes the procedure for protecting vSphere virtual machines provisioned or managed by vRealize Automation and associated with an IaaS machine blueprint. This includes machines provisioned by a user requesting an XaaS blueprint that calls a machine blueprint for provisioning. Machines or other resources provisioned by using XaaS directly do not have associated reservations, therefore this step is not necessary.