Recently there was an article published on the Microsoft Virtualization Blog which compared Hyper-V’s High Availability/Quick Migration capabilities to VMware’s VMotion. (VMblog pointed me towards the article) In the second article the writer responds on a large amount of reactions he had regarding VMotion being superior:
After my last blog I received almost two dozen email telling me that VMotion was far superior for unplanned host downtime and that it was a much better HA solution because it could live migrate virtual machines. I’ve heard this fallacy espoused for many years and, folks, this simply isn’t the case.
In the case of unplanned downtime, VMotion can’t live migrate because there is no warning. Instead you must have VMware HA configured and the best it can do is restart the affected virtual machines on other nodes which is the same as what is provided with Windows Server 2008 Hyper-V and Failover Clustering.
I can imagine why people reacted, in the first post the writer only mentioned VMotion. For unplanned downtime VMware doesn’t use VMotion because when it’s unplanned the VM’s get cutoff and will be restarted on another host with the use of HA(VMware High Availability). There’s no need for a migration when a VM is powered off.
Indeed Microsoft can do the same with the use of Clustering. But can you live migrate virtual machines when a server needs maintenance? No, at this moment that’s not possible. In other words, you will have to wait for a suitable moment… planned downtime, probably after business hours. But in a 24×7 environment will there ever be a suitable moment? Even when your business isn’t 24×7, if there’s a possible hardware failure would you want to wait? But when you have a 8:1 consolidation ratio you probably will not be the most popular system engineer when “quick migrating” the file server or the mail server especially when these VM’s have a lot of RAM assigned.
Besides that, with the upcoming new product, Continuous Availability, even unplanned downtime will not crash your VM. CA will constantly mirror your VM to another host, like a continous VMotion I guess, and when the active host fails the standby host will become active. In other words, no unplanned downtime anymore.