Hi,
We are starting an upgrade project for our SCVMM and Hyper-V infrastructure and I'm posting this in the hope someone has experienced a similar situation and has some advice. Our current environment is:
Six Windows Server 2012 R2 Hyper-V Clusters, three nodes in each cluster.
Two sites, each site has 3 clusters with Hyper-V replication used to replica VMs from main site to other site. This other site would be used in a DR fail-over situation.
A single physical SCVMM 2012 R2 Server, managing all clusters. This server is based in the main site.
The main production cluster is being replaced with new hardware and the original plan was to go with Windows Server 2019 Datacenter, also replacing the SCVMM server with a new physical SCVMM 2019 Server. This plan has changed now as SCVMM 2019 will not support the remaining clusters as they are 2012 R2 (SCVMM 2019 will only support 2016 or 2019 nodes). We are OK with this but is it a good idea to stick with a single SCVMM Server?.
I had read the Microsoft SCVMM High Availability documentation but this seems overkill for our environment. If there is an outage with the SCVMM node then I'm sure this will be an acceptable risk as it will not affect the running nodes and VMs. It was mentioned that we could run the SCVMM Server as a VM on the new cluster but I'm more comfortable with having this as a separate physical server. This Server will be backed up so worst case scenario for a failure of SCVMM is a restore of SCVMM but a DR fail-over to the 2nd site would leave use without a SCVMM server.
Any feedback on the above is welcome.
thanks