"Failover storage" for redundancy question

NWMrTIm

Cadet
Joined
Dec 5, 2023
Messages
3
Hey all, so I have an experiment I'm working on - perhaps even to develop a proof of concept.

I am wondering about having "failover storage" behind HA cluster VM Host nodes. Traditionally, HA clustering is 2 servers with a shared storage (and that in some ways is a single point of failure). My thinking is: if you can have failover virtual machine hosts, why not have failover storage?

I think this could be useful in two scenarios. First being 2 VHosts and 2 Storage units at one location, and the other being two different locations each having a VHost and a storage unit. I did some research on this and some articles say I need TrueNAS Scale for that kind of setup. Others have said: "oh, you can use TrueNAS core and r-sync it". Some feedback I've gotten ask for more design specifics, etc.

For now I'm just setting up a couple Hypervisor machines, and a couple NASs just to see if it functions. So could I have something like two NAS in the same pool that are "mirrored" for lack of a better term? and like if one NAS gets rebooted / unplugged / loses network or goes offline for some reason, the other NAS will continue to provide the HA cluster storage to both the VHosts? I'm just wondering it this is possible / feasible.

If there are already threads on this subject, I'd be happy to read them.

Thanks.
 
Top