In NLB all hosts work together to provide maximum performance, in Failover Cluster all nodes work together to provide redundancy. First I thought that Failover Cluster only 1 host works and if it fails the other host take over but now I see that all hosts work like NLB but if 1 host fail others take the load of the failed one. I am seeing that both technologies are very similar and therefore cannot distinguish the difference between them, can someone clarify please

Hi,

You can implement guest level NLB for application in Hyper V. For example, Web server. The traffic will load balance between both VM (if you configure NLB).

Failover cluster for Hyper V, you can implement host level or guest level failover cluster.

In Host level for Hyper V, you can configure Clustered shared volume. Both nodes can be active/active to provide redundancy.

In guest VM:- failover cluster, is for application. Most of times is Active/passive.

Other product such as SQL and Exchange is using active/passive clustering.

Can we use NLB to load balance the load of multiple Hyper-V servers in failover cluster of server 2012 which can automatically migrate VMs when one node is heavily loaded to the node which is less loaded?

NLB is historically used for web servers (the most common application server that implements it).

It is multiple machines responding to the incoming traffic on a single DNS namespace.

It is one deployment option of failover clustering.

Hyper-V and failover clustering together operate like a bunch of active-passive cluster nodes, all keeping the application alive.  However in this case the application runs in a container called a virtual machine (not installed on the OS).  And a virtual machine can only run on one physical machine at a time.

Thus failover clustering keeps the VM running, anywhere in the cluster.

You can user NLB to cluster VMs together.

You can implement SCVMM or other tools to rebalance your VMs based on load.