1.66.3 was released today , and I was trying to update it on a node that has auto-updates enabled. I went to the admin console and hit the "start update" button and received the notification that it was up and running.
Looking at the node in question I see from the journal:
May 15 20:53:39 appland tailscaled[623]: c2n: update command failed: exit status 1, output: Failed to start transient service unit: Connection reset by peer
May 15 20:53:39 appland systemd[1]: tailscaled.service: Got notification message from PID 3462, but reception only permitted for main PID 623
May 15 20:53:39 appland systemd[1]: tailscaled.service: Got notification message from PID 3462, but reception only permitted for main PID 623
May 15 20:53:39 appland tailscaled[623]: c2n: running "systemd-run --wait --pipe --collect /usr/bin/tailscale update --yes"
May 15 20:53:38 appland tailscaled[623]: c2n: GET /update received
May 15 20:53:38 appland tailscaled[623]: c2n: POST /update received
The update doesn't complete and dnf update does show it as still available:
sudo dnf update tailscale
Last metadata expiration check: 0:14:44 ago on Wed 15 May 2024 10:12:07 PM EDT.
Dependencies resolved.
Package Architecture Version Repository Size
Upgrading:
tailscale aarch64 1.66.3-1 tailscale-stable 25 M
Transaction Summary
Upgrade 1 Package
Don't see a similar issue, so thought I'd raise this one.
Steps to reproduce
See the upgrade available arrow in Admin Console
Click start upgrade
Upgrade says successful, but still doesn't apply the upgrade
Are there any recent changes that introduced the issue?
N/A
OS
Linux
OS version
Fedora 40
Tailscale version
1.66.1
Other software
This system runs nftables with a very basic firewall:
sudo cat /etc/sysconfig/nftables.conf
This is our tailscale exit node firewall that will
reside in another location other than our own
don't trust anything coming in on the local network
that we haven't established
but always allow incoming tailscale connections
table inet firewall {
chain incoming {
type filter hook input priority 0; policy accept;
ct state vmap { invalid : drop, established : accept, related : accept }
iifname {end0, wlan0} drop
}
}
What is the issue?
1.66.3 was released today , and I was trying to update it on a node that has auto-updates enabled. I went to the admin console and hit the "start update" button and received the notification that it was up and running.
Looking at the node in question I see from the journal:
The update doesn't complete and dnf update does show it as still available:
Don't see a similar issue, so thought I'd raise this one.
Steps to reproduce
Are there any recent changes that introduced the issue?
N/A
OS
Linux
OS version
Fedora 40
Tailscale version
1.66.1
Other software
This system runs nftables with a very basic firewall: sudo cat /etc/sysconfig/nftables.conf
This is our tailscale exit node firewall that will
reside in another location other than our own
don't trust anything coming in on the local network
that we haven't established
but always allow incoming tailscale connections
table inet firewall { chain incoming { type filter hook input priority 0; policy accept; ct state vmap { invalid : drop, established : accept, related : accept } iifname {end0, wlan0} drop } }
Bug report
BUG-4cdf818b0f9a73b8744cf16b6c8d23574f592b71d9cd3d1759d417d0dddac21d-20240516023044Z-2d19ef9f795e6ec1