This will not be a fun retrospective (the status updates are brutal and clear) -- and I've only been using Coveralls for open source projects for years.
There is a chance — depending on how stubborn the cloud infrastructure provider is (and both Google and AWS can be very stubborn because they don't like admitting that they just might possibly be wrong; I have no experience with Azure but imagine its the same) — that Coveralls will be unable to recover from this because rebuilding the infrastructure from zero on a different provider is going to be hard even if there is 100% IaC and even if there's a solid database backup outside of that provider that's accessible.
I hope they can, because they're a reasonable service and provide a lot of value to open source projects for coverage measurement.
That said, I have seen a number of reliable paths to getting extremely slow responses and eventually 500s from the coveralls servers while trying to look at the coverage details. It has felt like there's been a slow decline in Coveralls server quality because of that. (I've never really reported any of these because (a) I hoped that they were seeing these in their logs, metrics, or notifications and (b) I'm not a paying customer and it's easy for me to `open cover/excoveralls.html` or the equivalent.)
(I have tried the major alternative, codecov.io, in the past, but it's been a long time and I find it disappointing that they appear not to keep their example repos / documentation up to date.)
