Skip to main content
BlueXP disaster recovery

Create a resource group to organize VMs together

Contributors amgrissino
PDFs

After you’ve added vCenter sites, you might want to create resource groups that group VMs together. Resource groups enable you to organize a set of dependent VMs into logical groups that meet your requirements. For example, you might group VMs associated with one application or you might group applications that have similar tiers. As another example, groups could contain delayed boot orders that can be run upon recovery.

You can create resource groups using the following methods:

  • From the Resource groups tab

  • While you’re creating a disaster recovery or replication plan. If you have a lot of VMs hosted by a source vCenter cluster, it might be easier for you to create the resource groups while you're creating the replication plan. For instructions on creating resource groups while you're creating a replication plan, see Create a replication plan.

Tip Each resource group can include one or more VMs. The VMs will power on based on the sequence in which you include them in the replication plan. You can change the order by dragging the VMs up or down the resource group list.
About resource groups

Resource groups enable you to combine VMs that are related operationally and need to be protected as a single unit.

For example, a point-of-sale application might consist of several VMs hosting databases, several VMs hosting business logic rules management, and several VMs acting as a webserver-based storefront. It might be beneficial to manage the entire application’s availability with a single protection process by placing these VMs in a single resource group.

With resource groups set up, you can apply the replication plan rules for proper VM startup order, network connection, and more to ensure proper recovery of all VMs required for the application.

How does it work?

BlueXP disaster recovery protects VMs by replicating the underlying ONTAP volumes and LUNs hosting the VMs in the resource group. To do this, the system queries vCenter for the name of each data store hosting VMs in a resource group. BlueXP disaster recovery then identifies the source ONTAP volume or LUN hosting that data store. All protection is performed at the ONTAP volume level using SnapMirror replication.

If VMs in the resource group are hosted on different data stores, BlueXP disaster recovery uses one of the following methods to create a data-consistent snapshot of the ONTAP volumes or LUNs.

Relative location of FlexVol volumes Snapshot replica process

Multiple data stores - FlexVol volumes in the same SVM

  • ONTAP consistency group created

  • Snapshots of the consistency group taken

  • Volume-scoped SnapMirror replication performed

Multiple data stores - FlexVol volumes in multiple SVMs

  • ONTAP API: cg_start. Quieces all volumes so snapshots can be taken and initiates volume-scoped snapshots of all resource group volumes.

  • ONTAP API: cg_end. Resumes I/O on all volumes and enables volume-scoped SnapMirror replication after snapshots are taken.

When you are creating resource groups, consider the following issues:

  • A single data store should not host VMs protected by more than one replication plan.

  • Do not host protected and unprotected VMs on the same data store. If protected and unprotected VMs are hosted on the same data store, the following issues could arise:

    • Because BlueXP disaster recovery uses SnapMirror and the system replicates entire ONTAP volumes, the used capacity of that volume is used for licensing considerations. In this case, the volume space consumed by both protected and unprotected VMs would be included in this calculation.

    • If the resource group and its associated data stores need to be failed over to the disaster recovery site, any unprotected VMs (VMs not part of the resource group, but hosted on the ONTAP volume) will no longer exist on the source site from the failover process, resulting in failure of unprotected VMs at the source site. Also, BlueXP disaster recovery will not start those unprotected VMs at the failover vCenter site.

  • To have a VM protected, it must be included in a resource group.

BEST PRACTICE: Organize your VMs before deploying BlueXP disaster recovery to minimize “datastore sprawl.” Place VMs that need protection on a subset of data stores and place VMs that are not going to be protected on a different subset of data stores. Ensure that the VMs on any given data store are not protected by different replication plans.

Steps

  1. From the BlueXP disaster recovery menu, select Resource groups.

  2. Select Add.

  3. Enter a name for the resource group.

  4. Select the source vCenter cluster where the VMs are located.

  5. On the left side of the Add resource groups page, select the VMs that you want to protect.

    Screenshot showing the Add resource groups page

  6. Optionally, change the order of the VMs on the right by dragging each VM up or down the list. The VMs will power on based on the sequence in which you include them.

  7. Select Add.