I've been assigned to come up with a solution to best fit our organization needs. Basically, we currently have a single server running few bunch of VMs on one organization. And the need now is to segregate the network and server for a new branding office.
Our branding office will have 60 - 70 users in count within the next 5 years, according to the management. Along with that, All I'm looking for is 1 - 4 VMs in place and that's about it as the corporate doesn't have much things to run but domain authentication and DATA read/write. Based on that I was planning on something similar to below image;
This design is just an assumption as I'm not bothered to import/export VMs between the host manually (as backups) since the business can expect 2 - 3 hours of down time, in-case of an unavoidable circumstance (such a critical hardware failure, which rarely occurs). I could also go about and do a manual fail-over cluster but it doesn't really require for us IMHO. Correct me if I'm wrong!
However, below is what I'm looking at, based on the image shared;
- Buy two new identical servers with a good performance (Phys. CPUs, RAMs, Disks/Storage... etc)
- Install & Balance VMs between the hosts and do a replication to each other
- Implement DFSR (Distributed File System Replication) only for replicating file system (technically mount only one site to the users while leaving the other site unmounted, but the replication will happen, possible??)
What concerns me doing these;
- If DFSR is having both UNCs. Then, what sort of a risk I'm looking at (my colleague said DFS is good when its working but if there's power failure or unusual server causes, replication might have issues)
More updated version of the image;