But really, this is the eevblog forum, not a bank. How much effort and expense should I really go to ensure for every contingency?
As someone else said, thank you for the sanity.
I have CNC machines in our factory, and because they are big and expensive, people don't buy them to sit around doing nothing waiting for potential work. So the CNC machines are usually running a lot. Well, one went down - turned out a seal got compromised and coolant seeped in and destroyed an optical encoder. A day to diagnose the problem, a day or two to get a replacement part, and a day to R&R the machine and get it back up and running.
We were running some parts for a group buy for customers, and when told that we would be delayed by a few days, some of the responses were borderline comical. Some suggested we weren't a serious business because we obviously had no contingency plan. Some flamed us for having a single point of failure. Most "got it" and realized that shit happens.
Point being - whenever you talk technical with a bunch of tech guys, there will always be folks who will happily spend your money creating their dream system.. but those people 1) aren't signing the check and 2) aren't seeing all the facts like revenues/expenses and all the other factors that go into these decisions, so their opinions lack context and relevance.
In short, Dave, don't worry about the downtime... I got more work done today than usual, and I doubt you lost any users because of some hours of downtime on a FREE resource that delivers so much for that free price.