From time to time I ponder about all the terrible things that could happen: what can go wrong will probably go wrong so it helps to be (kinda) prepared.
As a single owner (of also a monitoring SaaS) I am currently putting out a fire where my primary datacenter died. I have suffered data loss even with precautions I had in place, one server lost all filesystems which took out my git repos.
I have backups, I have clones. I've still been in partial outage for 4 days and will be fully up tomorrow when I literally drive my servers to a new DC. Surprisingly I have slept 8 hours every night and I'm not worried. I've been in contact with my customers and provided solutions to keep them alive. If they leave they're going to leave, nothing I can do. I am looking to make sure everything is built uniformly (the server that died was the last of an old build process) and invest in scaling to the cloud in a bit more efficient and orderly manner.
