Not entirely shure where I’m taking a wrong turn here..

  • I’m provisioning a throw-away server using an emphemeral key
  • the server is tagged as “server” and “ephemeral”
  • I can see it as ephemeral in my Tailscale admin panel
  • I’m running tailscaled with “—state=mem”
  • I’m even doing an explicit “tailscale logout” on shutdown

Yet.. the instance sticks around in my Tailscale control panel for hours if not longer and blocks the hostname during that time.

Anyone in the same situation?

#tailscale #ephemeral #devbox #help

@heals I just tried to reproduce this, and it's behaving as expected in my lab. If you're able to get a `tailscale bugreport` right before the `tailscale logout` and send me the code that returns - either here or [email protected] I can dig in a bit further.

@jws thanks for trying to replicate my issue! I'll see if I can convince my terraform setup to get me a bugreport number :)

I'm entirely open to the possibility that it's just an issue with how terraform or scaleway destroys the server resource instead of an issue with Tailscale - neither bit of this is easy to debug sadly <.<