Anterovium - Fotolia

Can VMware's SRM do a failback?

VMware says Site Recovery Manager does failbacks and failovers, but one SRM expert says it's not quite the case.

How does a failback work in VMware's Site Recovery Manager?

VMware designed Site Recovery Manager (SRM) to minimize downtime for enterprises that need to get VMs back into production when a disaster strikes.

Using either vSphere Replication or a third-party storage-array replication product, VMware's SRM automates the process of synchronizing data between the primary and backup data centers. VMware recommends vSphere replication for companies with a relatively small infrastructure; bigger companies are advised to replicate VMs via storage.

In the event of an emergency, SRM automates the recovery process and brings up VMs in the order specified by the virtualization administrator.

According to one SRM expert, while VMware says SRM does a failback, that's not quite the case.

"There is no failback button. It doesn't quite work that way. They have what they call the 'reprotect,'" said Luke Huckaba, virtualization architect at Rackspace, during his presentation "Architecting a DR Solution -- Best Practices" at the Boston VMUG User Conference on June 24.

According to Huckaba, SRM reprotect begins the replication process from the backup site to the primary site and also "reverses your recovery plan."

"It knows that now where you're living is your protected site and the other site is the recovery site. So if you had some DR utility like the DR IP Customizer tool or any of that IP changing stuff, it reverses all that stuff as well."

Huckaba said that all the reconfiguration done by SRM enables a site "to go back in the other direction so you don't have to create another plan to go back. It's pretty slick, but it's not a failback."

Dig Deeper on Backing up VMware host servers and guest OSes