Ghaa, this is so frustrating. I've got a LAN, a #tailscale node in the LAN, another tailscale node elsewhere (remotely). I can ssh into the remote tailscale host from the local host that has the tailscale daemon running, but not from other machines in the LAN.

I have another LAN, very similar to the first LAN. I can ssh into the remote tailscale host just fine from any host in that second LAN.

The #headscale policies are the same for both LANs. I'm sure something is different, but after many hours trying to debug I can't find it.

Tcpdump shows traffic is going out on the tailscale interface on the local host that has the tailscale daemon running in the first LAN, but it's not coming out on the other end.

I'm sure it is some little stupid thing I'm not seeing, but ghaaaaaaaaaaa.

Any #verizon #fios customers here I can compare my experience with? I was downloading a large game from #steam and on my 500Mbps channel I was pulling around 250Mbps, and I was too impatient. Then I thought - hey, I'm doing it via my #Tailscale exit node on cheap VPS, it probably slows this down.

So I turn Tailscale off. And after a few seconds of speed drop, it stabilizes at about 150Mbps. Turn Tailscale back on - and I'm back at 250Mbps. Does #verizon throttle #steam???

Spend an hour looking at #windows "reparse points" because restic would not back up the parents locally synced Dropbox and for some reason the directory showed up in "dir" as such (or at least I interpreted it as such)... Then I looked into the (very sensible!) default exception list of https://github.com/kmwoley/restic-windows-backup ...

That script is great if you need to backup a windows PC... Works nicely with a #restic REST server on a NAS reachable via #tailscale, including #healthchecksio integration...