I’ve just gone and moved the whole application stack from one host to another to resolve this after connections from the first were being blocked, and now it’s happened on the second.
Not doing anything unusual and simply using Node Red to connect.
Yes, I saw that. It worked perfectly fine on both hosts and for quite some time (months to well over a year).
It strangely started working again on both last night. The hosts are with Hetzner (DE) and Bytemark (UK). Hetzner checked things out and confirmed zero filtering at their side. nmap showed the state as filter, so must have been dropped at TTN end, else en route. Former seems more likely, as other hosts in the same datacenter could connect fine…