hyper v replication what happen if host 1 crash?automatic replication? RRS feed

  • Question

  • I would like to know if i switch off my host 1 does the vm on host 2 that has been replicated will automatically start?

    What if my host 1 has a Domain controller in it and the replicated DC on host2?

    in fact i have 2 host with several virtual machines

    i start replication from host1 to host2 and vice versa from host2 to host 1.

    so if any of host fail i still can run the replication on the other host, but my question is , do i have to run it manually or it will automatically start?


    Wednesday, July 16, 2014 12:12 PM

All replies

  • Replication will not start up a VM automatically in case of host failure. Replication allows you to do a planned failover from the primary host to the secondary host.

    To do a planned failover:

    1. You must first shut down the VM(s) on the primary server, then

    While both the VMs on the primary and secondary hosts are off,

    2. On the Secondary host right click on the replicated VM, click Replication/Failover

    Hyper-V will start the VM on the secondary server.

    3. Next, you need to start Reverse Replication on the secondary server to replicate VM changes back to the primary during the failover period.

    When it's time to failback to the primary, you repeat the same steps, failing the VM from the secondary host back to the primary:

    Replication is meant for Disaster recovery, and ability to failover a workload from one data center to another. It is not meant as a hot-standby where a VM will stand up automatically in case of a failed Hyper-V host.

    If you need this functionality, you're looking for Failover Clustering not replication. A Failover cluster requires shared storage and will provide the functionality where VMs on a failed host will be automatically live migrated to other available hosts in the cluster with no down time.

    The last pieces you may consider is network load balancing and guest clustering. Failover clustering protects you against a Hyper-V host failure, but what if a VM blue screens or has internal failure not related to the host? Network load balancing and guest clustering provide some help there. For example, if you have a web server, you can setup a network load balancer (NLB) that will have a VIP (virtual IP) mapped to your web site name. Traffic intended for the load balanced web server will be balanced between the 2 VMs that should be setup with affinity rules to remain on 2 separate hosts. In case of 1 VM failure, the web site remains operational where all its traffic will be directed by the NLB to the operational VM. 

    Guest clustering is setting up a software cluster on top of a hardware cluster using shared storage that's presented directly to the VMs bypassing the hypervisor, or using shared VHDX. Note that as of the time of this note (8/8/2014), shared VHDX has some shortcomings including no ability to do snapshots or backup from the host (must use guest agent and backup the shared VHDX via old style backup)


    • Proposed as answer by Sam Boutros Friday, August 8, 2014 10:42 AM
    • Edited by Sam Boutros Friday, August 8, 2014 10:43 AM
    Friday, August 8, 2014 10:41 AM