Recently I had a power outage. Although I have a UPS, the performance is a bit patchy, as I often get an unclean shutdown (something to be sorted out later). I don't have any auto-restart on power restore set up; certainly in the past (quite a few years) the unRAID (7.0.1) server hasn't restarted when power was restored. Details are a little vague because I was travelling when the power went out, but I first found out the server was down when my phone, which was on Tailscale and uses the Tailscale exit node on the server, lost internet access. I disabled the exit node and internet access was OK again. I saw in the Tailscale app on the phone that Tailscale had lost connection with the server and some docker containers running on it (like Plex and Nextcloud). I was about to call home to see what had happened when I saw that suddenly I had access to the server again and everything was up, and in the Tailscale app I could see that the Tailscale exit node was up and the containers were connected.

Then last night it happened again (I was still away): lost Tailscale exit node and server seemed to be down briefly. But this time the "recovery" was not complete. The Tailscale node came back up and I got Tailscale access to the server but the Plex and Nextcloud containers didn't have Tailscale access. I was able to log in to the server and I could see that the two containers were running; they just didn't have their usual Tailscale connections. I confirmed that the Tailscale plugin was up to date.

My "evidence" of unclean shutdown and restart is that both times I got an email notification that a parity check had started because of unclean shutdown.

This is not a huge problem, but I'm wondering why this is happening and whether it's symptomatic of some misconfiguration on my part

tower-diagnostics-20250702-1041.zip

Edited July 2, 20251 yr by sonofdbn
Wrong attachment

Best to ask in the existing Tailscale plugin support thread.