Kevin Fogarty has started a lively discussion with his blog post called, “Is One of VMware’s Best Features a Really Bad Idea?,” in which he questions the sensibility of VMware’ real-time VMotion capability.
To continue the Iron Chef Virtualization metaphor I used the other day, VMware has one special ingredient that their competitors don’t: the ability to move a running VM from one physical host to another with almost no interruption in service (in contrast, Microsoft and Xen can move a VM, but it is a “cold” move that takes the VM offline during the move).
Fogarty posits that live VMotion capability is a risky feature – in fact, too risky for practical use in production. In the comments on the post, there are a number of folks that insist VMotion is practical in production, and that they use it every day for load balancing.
From what I know about VMotion, it is a very persnickety feature with all sorts of requirements to make it work but it is still very cool when it does work.
Today, an argument can be made that real-time VMotion is VMware’s single biggest competitive differentiator (from a technical capabilities perspective). My questions:
- Is VMotion enough to preserve VMware’s dominance in the market, or will enterprises find that Microsoft’s “cold” motion is sufficient?
- Will other virtualization vendors develop “hot” motion capabilities? If not, why not? If so, when?
What do you think – does this one ingredient make VMware’s recipe so much better that you’d rather fight than switch? Is “hot” motion too risky for your environment? Or is “hot” motion even important to you as a virtualization buyer?