The move to server virtualization usually incorporates high-availability technology to protect mission-critical virtual machines (VMs). But data center high availability (HA) certainly is not the end, according to Pierre Dorion, a data center practice director at Longview Systems in Denver.
Implementing hypervisors such as VMware ESX, Microsoft Hyper-V or Citrix XenServer and then adding software features such as VMware's Distributed Resource Scheduler (DRS) and VMotion may create a false sense of security -- where a data center manager believes that a virtualized data center is prepared for any eventuality, only to be caught completely by surprise.
Data center high-availability technology and plans must extend beyond virtualized servers to encompass networks and storage, along with other critical facilities such as power and cooling, Dorion said. If the loss of any single physical infrastructure component prevents your virtual infrastructure from being available to end users or to support business processes, you do not data center have high availability -- even if virtual servers still run, Dorion said.
The IT infrastructure is only as redundant as its weakest component.
"I tell my peers regularly that you can install airbags on a car with no brakes, but it is still a car with no brakes," Dorion said.
As always, testing is the best way to ensure undesirable results are addressed, or at least understood, before an unplanned outage takes place.
Be sure to extend your testing to facilities, not just to data center systems, Dorion said. It's not just a matter of what happens when a single server fails. Instead consider what happens when network or storage switches fail, when power fails, when the air-conditioning system fails and so on. This doesn't mean everything in an IT environment has to be highly available, but it does make sense to identify all single points of failure in servers, network, storage and facilities, and plan around them before an outage, Dorion said.
The Uptime Institute has developed an industry-accepted Tier Classification System to rate data center high availability, which underscores the fact that high availability technology goes way beyond the ability to fail over computer systems.
About the Author
Stephen J. Bigelow, a senior features writer at TechTarget, has more than 15 years of technical writing experience in the technology industry. He 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.