Machines just stopped being accessible?

I’m running tailscale on a oracle cloud CentOS and multiple clients - Ubuntu & Mac. So i was running 2 HTTP servers, on port 8080 and 8083 and they were reachable from other machines just fine.

Then i added a new server on port 8085, and it wasn’t accessible. The older servers were still accessible. I restarted the servers on the other ports, and nothing is accessible now. ping and the tailscale ping work, but otherwise curl says no route to host or connection refused. Even the tailscale services page shows the service running on port 8080, but i cant connect to it.

#on oracle server
[opc@nopearm ~]$ sudo python3 -m http.server 8080

#from ubuntu
nope@ubuntu:~$ curl nopearm:8080
curl: (7) Failed to connect to oracle port 8080: No route to host

#from macbook
nope@macbook:~$ curl nopearm:8080
curl: (7) Failed to connect to oracle port 8080: Connection refused

I can still run servers on other machines, but the oracle box seems to have gotten stuck in a weird state… how do i get it working again?

I had a similar experience where all my three Home Assistant devices became unreachable. I found out that all of them were logged out from Tailscale. I needed to log back in via Home Assistant Tailscale addon and now they are reachable again.

I also just learned that I can disable key expiry in Tailscale Admin panel for these machines that are a bit problematic to renew the key.

Thanks.It was logged in, i forgot how i fixed it now but its fixed :slight_smile: