I've considered such a setup myself and decided that I also wanted redundancy at the hardware/node level. If my storage server goes down, then everything is dow

I've considered such a setup myself and decided that I also wanted redundancy at the hardware/node level. If my storage server goes down, then everything is down until it's fixed. I think the person who made this setup had the same considerations