4

Forgive me if I am missing something obvious but I am very new to servers and failover clustering. I haven't found a question that quite answers what I am looking for yet.

My goal is to make a VM run 100% (or near enough) of the time using two or more servers in a cluster. This is for an access control system hence why it needs to be up all the time and failback seamlessly, so client workstations may access this. From what I have read is exactly what failover cluster is designed to do and should be perfect for this.

I have 3 physical servers, 2 of them clustered and 1 the Domain Controller. Using iSCSI i have installed the VM in shared storage and can run it and connect to it via RDP on any of the nodes or the DC but when one of the nodes is disconnected. Each time this happens i see in the setting of the VM the virtual switch is not avaliable to use and have to make a new one so it will be avaliable for conenction.

Am I missing a hardware component? Is what I am trying to acheive actually possible?

Hardware Specifications. I am aware that these servers are very far apart in term of hardware but the application I am needing to run on the Vm is not very demanding.

Domain Controller,

Xeon E3-1230 V2, 16GB DDR3, 1TB HDD,

Node1,

Intel i9-10900, 32GB DDR4, Server 2019, 500GB SSD,

Node 2

Intel i5-9400F, 16GB DDR4, GTX 1660, 500GB SSD

Virtual Machine 4 cpu cores 8GB RAM 60GB Max hard drive size

Thank you in advance if anyone has any pointers for this.

Meredith
  • 41
  • 3

1 Answers1

4

You need to create Failover Cluster and use iSCSI shared storage to create Cluster Shared Volume (CSV). CSV will be used as a storage for your VMs. https://docs.microsoft.com/en-us/windows-server/failover-clustering/failover-cluster-csvs

The following guide covers the configuration process: https://www.starwindsoftware.com/resource-library/starwind-virtual-san-for-hyper-v-2-node-hyperconverged-scenario-with-windows-server-2016/

As for DC, I would recommend you to place it outside of the cluster. You should have at least 2 DC VMs, each running on a separate node from local storage.

Stuka
  • 5,445
  • 14
  • 13
  • 2
    Thank you for that link and the answer! I was able to get the cluster set up with a VM switching between the nodes. The DC will be un-clustered and is a physical server too. Is there a reason I would want 2 DC VMs? As the VM I am planning to run will contain it's own server for a proprietary access control application that client workstations need to see. That VM just needs to failover between the nodes and resume services once booted, nothing more. I'm also guessing that star wind would be required for it to qualify as fault tolerant rather than highly available? – Meredith Jun 14 '22 at 11:44
  • 3
    StarWind VSAN creates fault tolerant shared storage for the VMs. Failover Cluster handles high availability and failover of VMs. As for DC, for homelab it is ok to have one DC, but for production environment it is recommended to have at least two, so that DC is available for most of the time. – Stuka Jun 16 '22 at 21:39