In my first post about Bit Mountain I should have clarified that it's a research project, paid for by my employer, designed to find the most cost effective way to reliably store petabytes. Bit Mountain isn't serious yet, it's just the yardstick by which I measure all of the storage vendors. I need peer review on this project, so I thought it would be a good idea to start discussing it. I would get a lot more peer review if it were open source, but that decision is up to my employer.

My hope is that someone will say, "Stop! Doing it that way will lead to tremendous administration headaches!" and sell us something much better. But the outlook for that doesn't look good right now. Everyone just wants to sell us continuous-power SANs, but none of them can tolerate more than two lost drives. That just won't do if you're going to power off the system, and I think local storage guru Josh Coates would agree.