What happens when you take OES Linux clustering and add XEN virtualization to the mix? The combination can be quite powerful when creating high availability for services. In the traditional cluster, services are made highly available. Should it fail on one node, the cluster starts it up on another node. Or if a server needs to have maintenance done on it, the cluster services are moved to other nodes. Moving the clustered service generally means shutting it down on one node, and starting it up on another. But what if the service was an XEN virtualized OS? Instead of having to shut down a service, XEN can be told to live-migrate the virtualized OS over to a new node -- resulting in virtually no noticeable outage to any connected clients. One of the BrainShare sessions I attended today went through a demonstration where they cluster-enabled a XEN virtualized OS, and moved it around the cluster. Any persistent connection to the OS was maintained as it moved from one node to another. This was all done with the currently shipping OES2 product.