VM-centric Disaster Recovery Done Right

Virtualization is setting the foundation for transforming enterprise datacenters into private clouds. In this flexible, resource pooled, dynamic environment, how do you maintain performance and availability SLAs, while providing all of the other benefits like self-service provisioning and resource elasticity? Delivering 100 percent service availability means having a proactive disaster recovery (DR) solution that is designed from the ground up for virtual workloads. Not surprisingly, this is easier said than done.

Traditional DR/Business Continuity solutions are based on simplistic array-level block replication and therefore deliver on only storage level DR.  To support overall application level DR one is then forced to cobble together a solution on top of the existing array-level block replication technology. While minimally functional for small environments, this approach doesn’t scale. Before you know it, this approach becomes an end unto itself.

Nutanix DR is converged DR in the truest sense, providing application level DR for both storage and compute and has been designed to provide service level continuity. Nutanix does this by providing VM-aware DR from the ground up. The concepts are simple, yet powerful:


1. A Nutanix Protection Domain is a grouping of VMs that the admin wants to protect. Each Protection Domain can be assigned a Replication Schedule. Each Protection Domain can replicate to any number of remote Nutanix clusters.

2. Nutanix replication schedules are very flexible. Not only is there support to do the usual hourly, daily, weekly schedules, but the admin can also specify an arbitrary frequency, such as:

  •  Take n number of snapshots every hour, but replicate every other Y snapshot to all remotes.
  •  Take m number of snapshots every hour, replicate Y snapshots per hour to remote1 and z snapshots per hour to remote2.

This allows custom SLAs per VM, or group of VMs – without having to move them to different datastores and thereby enables enhanced flexibility to manage replication, taking into account compute, storage, networking and application requirements.


3. In scenarios where application dependencies require that all related VM and file entities have a consistent cut of their internal state, admins might want to group VMs/files together so that they are snapshotted consistently. Nutanix DR groups these VMs/files in a Consistency Group. A Consistency Group guarantees that a snapshot across all its entities is consistent.  A Protection Domain can have multiple Consistency Groups.

The fact that Nutanix DR is VM-aware gives the admin maximum control to manage how network bandwidth is used. For example, critical VMs can be set-up to replicate at a higher frequency, and less important VMs can be replicated less often. This is in contrast to traditional storage arrays that force replication at a LUN level. As LUNs can often be on the order of TBs, such coarse-grained replication can overwhelm site-to-site WAN bandwidth resources.


4. During failover the Nutanix admin has the flexibility to failover the entire Protection Domain or subset of VMs. This is another example of how Nutanix VM-aware DR is superior to other DR products in the market. In the case of DR products from storage vendors, replication is at the LUN/volume level, which means it is an all-or-nothing, either you have to failover all the files on the LUN or none. There is no fine- grained capability.


In a nutshell, traditional DR is no longer sufficient for modern datacenters. End users expect 100 percent application availability and reliable performance SLAs. Nutanix DR provides holistic, true application level DR enabling IT teams to take a more proactive approach to managing outages.