Manage Learn to apply best practices and optimize your operations.

Virtual disaster recovery and server backup case study

Kroll Factual Data went with a full virtual disaster recovery solution when its virtual server backup software and tape libraries weren't cutting it anymore.

For companies tasked with supplying up-to-the-moment financial or customer information, virtual disaster recovery (DR) and server backup are critical parts of everyday business operations.

The need for constant virtual server backup and data protection is even more pronounced for highly virtualized organizations like the facility at Kroll Factual Data, an information services business located in Loveland, Colo. As an early adopter of server virtualization, Kroll Factual Data now touts a Microsoft Hyper-V environment supporting about 1,700 virtual machines (VMs) across disaster recovery, development, quality assurance (QA) and production groups.

More on virtualization backup and DR
Virtual data backup best practices

VM backup problems

How to configure virtual server backup

How to optimize server virtualization backup

Virtual data protection and recovery checklist

Virtual disaster recovery case study

The virtualized data center has performed solidly for Kroll Factual Data, but the traditional use of Symantec Corp.'s Backup Exec and tape libraries was clearly no longer adequate for virtual server backup (22 servers and 650 VMs in the production area alone).

"You would back up critical systems and hope the tape backups worked and do testing here and there to make certain that you had some reliability in your processes," said Christopher M. Steffen, the principal technical architect at Kroll Factual Data.

The move to virtual disaster recovery
Steffen said he still uses tape on about 10% of the environment but has shifted VM backups to snapshots stored on a SAN and continuously replicated to a DR site.

Steffen said that VMs are frequently updated, so snapshots are constantly reloaded, updated and cycled through Kroll Factual Data's development and QA groups before assuming a role as new golden images. Consequently, backups are regularly verified and tested as a regular part of business workflow.

"We make continuous copies of the golden image files of all our virtual machines," Steffen said, pointing to virtual disaster recovery and business continuity as the immediate beneficiaries.

New VMs are deployed quickly to new servers using those established images.

"We also migrate our production changes through virtual file, not through application updates," Steffen said.

Optimizing virtual server backup
For Kroll Factual Data, this involves making application changes to a VM in the development group, testing those changes in the QA group, then staging the updated VM and finally passing that updated VM to the production environment. This process avoids the compatibility and performance issues that often plague regular application development workflow.

"If it worked in your dev environment, it for sure is going to work in production, because you're talking about the same [virtual] machine," Steffen said.

Even images not updated by the development team are patched and updated at least monthly. Steffen said that the current virtual server backup process is much simpler than traditional tape and has delivered tremendous confidence to the Kroll Factual Data organization.

"It literally is as simple as loading up the virtual machine, the VHD [Virtual Hard Disk] file, getting it up on the host, and having it go," he said. "It really has become that straightforward to us."

Dig Deeper on Disaster recovery, failover and high availability for virtual servers

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.