Probe had a tunnel error. It's now working properly again, but had to suspend it for several hours. What was the cause?
Found that the site was unreachable from three different stations in the same logical location, two kubernetes and one docker-compose, plus from one station on a different ISP, so highly unlikely to be a station-related issue.
Looked that the check result. The actual error is 'Unable to connect to host'
So here's all we know about the failure (and everything that we can possibly know from our side)