Tailscale unable to start

After upgrading tailscale to the 0.19.0, it stopped working completely.

Here are the relevant logs

 Add-on: Tailscale
 Zero config VPN for building secure networks
-----------------------------------------------------------
 Add-on version: 0.19.0
 You are running the latest version of this add-on.
 System: Debian GNU/Linux 11 (bullseye)  (amd64 / qemux86-64)
 Home Assistant Core: 2024.5.4
 Home Assistant Supervisor: 2024.05.1
-----------------------------------------------------------
 ...
Log level is set to TRACE
...
[10:32:18] INFO: Starting Tailscale web...
[10:32:18] INFO: Starting Tailscale...
[10:32:18] TRACE: bashio::addon.ip_address
[10:32:18] TRACE: bashio::config.has_value: userspace_networking
[10:32:18] TRACE: bashio::addons self addons.self.ip_address .ip_address // empty
[10:32:18] TRACE: bashio::cache.exists: addons.self.ip_address
[10:32:18] TRACE: bashio::config: userspace_networking
[10:32:18] TRACE: bashio::fs.file_exists: /tmp/.bashio/addons.self.ip_address.cache
[10:32:18] TRACE: bashio::addon.config
[10:32:18] INFO: Allow UDP GRO for forwarding on enp1s0
[10:32:18] TRACE: bashio::var.false: self
netlink error: bit name not found (offset 56)
netlink error: Not supported
[10:32:18] TRACE: bashio::cache.exists: addons.self.options.config
s6-rc: warning: unable to start service post-tailscaled: command exited 92
[10:32:18] TRACE: bashio::cache.exists: addons.self.info

My system:

root@ha:~# cat /etc/os-release
PRETTY_NAME="Debian GNU/Linux 11 (bullseye)"
NAME="Debian GNU/Linux"
VERSION_ID="11"
VERSION="11 (bullseye)"

root@ha:~# dpkg-query --list | grep homeassi
ii  homeassistant-supervised                1.4.1                            all          Home Assistant Supervised

root@ha:~# ha info
arch: amd64
channel: stable
docker: 26.1.3
features:
- reboot
- shutdown
- services
- network
- hostname
- timedate
- os_agent
- resolved
- disk
hassos: null
homeassistant: 2024.5.4
hostname: ha
logging: info
machine: qemux86-64
operating_system: Debian GNU/Linux 11 (bullseye)
state: running
supervisor: 2024.05.1
supported: false
supported_arch:
- amd64
- i386

It turns out the 0.19.0 is somewhat broken with my setup.

I found a way to revert this temporarily by renaming its docker image 0.19 to 0.18

Make sure the tailscale add-on is stopped. Then, do

docker pull ghcr.io/hassio-addons/tailscale/amd64:0.18.0
docker tag  ghcr.io/hassio-addons/tailscale/amd64:0.18.0 ghcr.io/hassio-addons/tailscale/amd64:0.19.0

Then restart the add-on.

Yes, I’ve seen this problem today. Luckily I’d checked “Create backup before update” enabled, so just restored that.

1 Like

Same, restoring from a backup made the trick. DAMN TAILSCALE!!

I have the same problem. I’ve also made a backup. Can you just restore the tailscale addon? Or do you need to run the whole HA restore?

Edit: you can just partially restore. Fixed now :grin:

I also noticed today that the (I think it was) 0.19.0 pending update has now disappeared, so the dev guy(s) are well aware of the issue.

I’m a bit of a novice with Linux in general, so I regularly backup the whole drive as well as using the backup in home assistant before I make any changes.

I didn’t realize about the backup. It is too late for me to use the backup now. :man_facepalming:

I just noticed that 0.19.1 was released. Given I was using a workaround using docker tagging, it overwritten it and still doesn’t work.

Same problem here. I also reverted back to the addon version 0.18, which still works (although in the tailscale web UI machine overview there is a warning that the used tailscale version 1.64.0 has security vulnerabilities).

Addon versions 0.19 and 0.19.1 both didn’t work.

I guess we should open an issue here.

1 Like

Addon version 0.20 is out. Did anyone try if the problem is fixed?

I’ve updated to 0.20 and its stopped working again.

OK, I’m still at 0.18. So maybe someone with 0.20 should report the problem with logs here. As the add-on is actively developed I think there is a good chance it will be fixed.