There is no one right way to implement a virtual server backup system, and a conventional tape approach can still be used -- even though there is little reason
Still, before you configure virtual server backup, consider some common elements. Virtual server backup always starts with storage to hold the virtual machines (VMs) captured from each server. The storage is often shared storage across a Fibre Channel or iSCSI SAN that is accessible to every server in a data center.
Shared storage is important because it allows rapid recovery of VMs to almost any other connected server. Organizations that are new to this type of virtual server backup may need to add considerable storage capacity to hold each VM dynamically, loading it into server memory so that it can run. This, in turn, increases storage costs but is usually offset by the benefits of greater VM availability.
Next, review your virtual servers and decide on the necessary level of protection for each VM. Critical VMs can receive continuous snapshot protection, although IT administrators might choose to take a VM snapshot backups of less-critical VMs every 30 to 60 minutes. Pick a recovery point objective, or RPO, that makes sense for each individual VM. A server evaluation should also weigh the implications of VM failover and load balancing in a virtual server backup and restoration scheme.
A key benefit of virtualization is the potential to restore VMs from storage onto other working servers without serious concern for the underlying hardware. This is a critical means of improving application availability, but each server requires the computing resources to support all the VMs operating on it.
Many IT professionals new to virtualization concepts can run server utilization close to 100%, only to discover that there is no failover capacity.
"One of the servers goes down, and there's nowhere to put those VMs," said Bill Kleyman, the director of technology at World Wide Fittings Corp., a manufacturer and distributor of steel hydraulic tube and fittings in Niles, Ill.
He added that virtualization software often allows administrators to pre-define how certain VMs will fail over to specific servers until a failed system can be repaired.
After they configure virtual server backup, IT administrators need to test and verify those systems to ensure their integrity and recoverability. In traditional backup environments, testing was extremely troublesome; the only way to truly verify a backup was to restore it to a live system and use it. Most IT administrators did not have the time or nerve to attempt a test restoration, and verification was often put off until absolutely necessary. But testing virtual server backup is dramatically different.
"One thing you want to test is live migration," Kleyman said. "Make sure you're able to migrate all your VMs from one physical server to another live without any downtime."
Live migration typically involves a technology such as VMware VMotion to move a VM between servers and verify proper operation.
A second test is to verify that a VM can be reloaded from its stored SAN image and resume operation without data loss. But for many organizations, the risks of restoring to live production servers are usually too great, and each VM can be restored to one or more virtual servers configured in a lab environment.
"Another test is that you're able to spin up a new virtual machine and load it from an image that you've saved, so make sure your virtual machine boots back up," Kleyman said.
An IT administrator should examine the VM to see that it loads and starts and should ensure that all applications in that VM is fully available with no data loss. A VM can then be offloaded, and another one tested. This type of process is transparent and does not have an impact on users. IT administrators can test a virtual server backup and practice restorations far more actively than in traditional environments.
About the author
Stephen J. Bigelow, a senior technology writer in the Data Center and Virtualization Group at TechTarget Inc., has more than 15 years of technical writing experience in the PC/ technology industry. He holds a bachelor of science in electrical engineering, along with CompTIA A+ and Network+ certifications, and has written hundreds of articles and more than 15 feature books on computer troubleshooting, including Bigelow's PC Hardware Desk Reference and Bigelow's PC Hardware Annoyances. Contact him at firstname.lastname@example.org.
This was first published in May 2010