The only solution to running your production systems if SAN dies - is to have another storage system (local, DAS, SAN, NAS). You could run from backups, but that will be slow. Here is how i would setup your systems:
- Virtual Veeam server with 4-8 vCPU and about 8GB RAM
- Install NL SAS drives in RAID 10 and configure
- schedule replication from SAN to local storage on server hourly (if your systems are fast enough, if not - then 2 hours or whatever schedule fits your environment)
- schedule daily backup to Promise (reverse incremental) and maybe a second to remote site via fiber. If you want to run only one job - then forward incremental to promise with active full on a weekend and rsync to remote site.
You want another proxy at remote site if you are replicating your systems to a server there (another virtual host) or if you have NAS for storage...
I recommend a physical backup server instead of virtual.
You can use additional virutal proxy server but better have the backup system on a standalone host that you can have immediate access to in case of problem with production system (storage, san netwrok, vm hosts).
Installing DAS disks on the esxi hosts in such environment is always a good idea IMHO, and you can use them in several ways for backup,
testing, and other uses.
I would go with RAID 5 in favor of capacity vs performance.
Regarding replica of the backup - how did you plan to achieve it?
If possible, I suggest running backup job directly over the Fiber, instead of replication the backup.
Create different backup jobs for local and remote backups.