Network Architecture with Hyper-V Clusters

To get very high uptime, servers need to have failover capability. If the primary server goes down for any reason, another one takes over for it. The failover machine has the same software and shares access to the primary server's data.

The traditional approach is to have a backup machine for each primary, but virtualization presents a more flexible and economical solution. Hyper-V clusters allow you to set up a failover cluster, with up to sixteen nodes (physical computers) that can all do the same work. The cluster can host up to a thousand virtual machines running Windows Server.

All the nodes can have access to Cluster Shared Volumes, which they can read and write. With this configuration, the failover node can restart the VM and use the same data that the failed node was using. The shared volume can reside on a disk drive or a Storage Area Network (SAN) device.

In addition, the Hyper-V cluster allows load balancing by migrating virtual machines from one node to another. Live migration is supported, allowing a VM to move between nodes without interrupting its operation. There's no limit to how many VMs can live-migrate at the same time.

For best results, all the nodes in a cluster should have identical or very similar hardware. At a minimum, the processors all need to come from the same manufacturer; live migration doesn't work between different brands. The BIOS of each node needs to support native virtualization. Each node needs multiple network cards to handle all the cluster's levels of network connection.

SafePoint IT can set up a cluster that will satisfy the highest standards for reliability. Please contact us for more information.