DO does not have a good support channel. My droplet died. I couldn't reach it, and neither could the world. There was no emergency button for getting help. Just send in an email. After a small time window, I had to just delete and rebuild my droplet. Two days or so later, they got back to me and because I deleted it, they couldn't debug it. Two days to get back to me on a dead, non-reachable droplet. I don't think that is acceptable for anything running in production.

But the whole point of treating cloud servers as cattle rather than pets is that when one dies, you can spin up another one, right? Ironically, that's one reason I prefer AWS over DO and the like, because AWS's EC2 auto-scaling and availability zones are great for this kind of resilience.