Oct 27
Virtualization with Distributed Power Management
Eric Inch
I enjoy learning, using and helping others through technology. This is my fourth year with C/D/H after many years of consulting for numerous small and mid-sized companies. I enjoy challenging projects and continual improvement in all areas. Most recently, I have been working to help grow the unified communications and virtualization practices at C/D/H.
When I’m not working, I enjoy spending time with my family. My girls keep me extremely busy but are always the highlight of my day.
For a more in-depth bio and a list of my areas of expertise, please visit http://www.cdh.com.
More about Eric
Articles by Eric Inch
This past week, Jason Sharp and I put on a presentation for C/D/H’s Annual Technology Briefing Series on the topic of virtualization. At a high level, we addressed numerous technologies and trends in this growing area. The one discussion we had with the attendees that still has me thinking was power consumption and resource requirements. I think this area really expands on the numerous benefits of virtualization and server consolidation.
The basis of this discussion was that technology is moving toward automated power management for servers. VMware has addressed this in their Distributed Power Management (DPM) feature. It is included in ESX 3.5, but is currently available only for testing purposes. VirtualCenter will continuously monitor resource requirements of currently running virtual machines and the power consumption of the current ESX hosts included in a cluster. Once the cluster can use fewer physical resources than allocated to run the virtual machines, the DRS cluster will consolidate virtual machines on fewer hosts and put other hosts into a standby mode. This effectively reduces the power needed to run the additional server(s). When the requirements increase, the hosts that are currently powered down are then brought back online to meet the necessary service levels.
Pretty cool, wouldn’t you say? You are running your virtual machines across all nodes in the cluster during the day when resources are highly utilized. During the night hours, the cluster automatically consolidates the VMs to fewer nodes and powers off additional servers. This seems to be an exciting extension of an already highly beneficial area of virtualization.