Skip to main content
Question

Looking for PC to PC DR deployment advice

  • November 17, 2025
  • 3 replies
  • 51 views

Hi all,

I have 2 sites with a cluster in each site. The workloads are distributed between each cluster. I currently have one Prism Central managing both clusters.

I have DR configured to a passive site hosted by a 3rd party in a datacenter in different region.

There are resources on each cluster available, so now I want to configure DR between the clusters. At least for the critical VM’s. This should allow me to failover the critical servers quickly if only one site goes down. I’ll use the passive site if both go down.

I’m assuming I will need to deploy another PC on the cluster that doesn’t currently have one. Ideally, I still want to manage both clusters from the primary PC in the co-lo facility. Is this possible? Or do I have to unregister the remote cluster from the current PC first then register it to the newly deployed PC then deal with managing each cluster from a different PC?

I also read that I should be able to deploy the new PC directly from PE via the “Deploy New Prism Central” option from the “Prism Central Registration” menu. Unfortunately, it’s not there.

I’m running pc.2024.3.1.9.

Any links to specific documentation or other help is appreciated.

Thanks!

Dan

3 replies

JeroenTielen
Forum|alt.badge.img+8
  • Vanguard
  • November 19, 2025

This is a design question. You can keep the same prism central as you are already using. But when the cluster/datacenter goes down where the PC is running you cannot do the DR. You first need to restore PC on the other cluster before you can do DR. Depending on the SLA's (RPO/RTO) this can be a thing. And restoring PC will not restore all configuration (eg Flow and templates are gone). 

 

If you do want to overcome a complete cluster/datacenter disaster where the PC is running then you need to install a PC on the other cluster as well and connect that cluster to that pc. Create availability zones between them and you are good to go. 

 

You can also run PC on the third DC (if that is possible though) and manage all cluster from that pc. If the third cluster/dc goes down the other two should be continue running and you can restore pc/cluster on third location without any failing SLA's.

 

So it all depends on the SLA's. ;)  


  • Author
  • Voyager
  • November 19, 2025

Hi Jeroen,

Thanks for the response.

Our RPO/RTO’s are in hours not minutes. I would rather not restore the PC on the other site if I can help it. Hence the reason I’m trying to deploy one on the other cluster.

You say I need to install a PC on the other cluster and connect that cluster to that PC. You also mention a third PC can be used to manage both clusters.

Are you suggesting my best option would be to like so:

    PC-1 (Cluster-A registered)
    PC-2 (Cluster-B registered)
    PC-3 (Cluster-A and Cluster-B registered)

I’d rather have this. Is it an option?

    PC-A (Cluster-A and Cluster-B registered)
    PC-B (no clusters registered).

Thanks again,

Dan
 


JeroenTielen
Forum|alt.badge.img+8
  • Vanguard
  • November 20, 2025

No that is not what I'm saying. That is just another method. So these are the DIFFERENT options/methods:

  1. Keep is as. And then restore PC on the other cluster when the “pc-hosted-cluster” goes down. That mean an additional RTO from 1 to 2 hours. 
  2. ClusterA-PCA and ClusterB-PCB. With availability zones. Then you have two prism centrals. When something goes wrong you can easily restore the virtual machines on the remaining clusters. For single management pane use Nutanix Central. 
  3. ClusterA and ClusterB againt Prism Central on a third site. Single management pane. And when clustera or clusterb goes down you can easily recover the vm's on the remaining cluster.