Tailscale workaround on OpenBSD
After a sleep-resume cycle i have often found the tailscale network in a non-working state on my laptop. Most of the time the ts network was unaccessible, but some times the clear. I have suspected a TS, a DNS or a WLAN AP reconnection issue, and investigated it a bit but it is time-consuming and a PITA to check what happens if a wlan AP leaves and an other one shows up repeatedly.
Now i am using a workaround since some time and i can happily report, this solved the issue for me. It won’t necessarily work for you, so consider this post as a notes for myself:
~> cat /etc/apm/resume
#!/bin/sh
sleep 20
rcctl restart tailscaled
~>
TLDR: restart the tailscale daemon after resume with 20s delay.