I posted on this earlier, but wanted to follow up with some posts by some smarter folks, Mike Laverick, Chris Wolf, and VirtualGeek have now all blogged on this, and the general consensus is that DPM is still “experimental” and unless there is a compelling reason, it may be best not to play with fire. My favorite quote “it would brave man who would rely on a magic packet to power on an ESX host. For me in the world corporate datacenters – the only thing that powers on or off a physical host is human being authorized by change-management request…”
Read the posts and decide for yourself:
Mike Laverick – http://www.rtfm-ed.co.uk/?p=691
Chris Wolf – http://www.chriswolf.com/?p=203
Virtual Geek – http://virtualgeek.typepad.com/virtual_geek/2008/12/does-vmware-dpm-shorten-esx-server-lifespan.html