[Bug?] Socket-proxy suddenly quits

[okynnor]

What’s the bug?

socket proxy quit with code 0 - Docker

Description

When the main Tugtainer is set to run, the socket proxy quits. This has been happening every other day. It ran by checking and downloading new container images and updated them one day. By the second day, it crashed.

Below is the log:

[ls.io-init] done.
[NOTICE] (1) : Initializing new worker (11)
[NOTICE] (1) : Loading success.
[NOTICE] (1) : haproxy version is 3.2.8-9200f39
[WARNING] (1) : Exiting Master process...
[ALERT] (1) : Current worker (11) exited with code 143 (Terminated)
[WARNING] (1) : All workers exited. Exiting... (0)

I suspect that this is only in Tugtainer because it doesn't happen in Watchtower. I link like this because both are on running on the same Docker host (Watchtower is inactive though. So they are not running concurrently with Tugtainer.

I notice that while I have checked to have Tugtainer check for updates, it seems that it is able to find an update for the socket and downloads it. It just doesn't update it. I believe, it is due to the "Protected" setting? I just want to let you know ...not sure if this is related to this frequent crash of the socket or not.

Environment:
Docker version 28.1.1, build 4eba377

Expected behaviour

It should run the cron job - namely to check and to update containers at 15:00h daily

To reproduce

Let the socket run for 2 days with a cron job set ...for me it was at 15:00h daily

The remote socket is unaffected and runs...meaning that it loads into tugtainer main webUI. Yes, the main tugtainer WebUI runs ...just the socket crashes.

Platform

Docker version 28.1.1, build 4eba377
Running inside a VM

Versions

Tugtainer v.1.12.2

Relevant log output

Extra

No response

[Quenary]

Hello,
I'd say tugtainer most likely has nothing to do with this. My socket-proxy has been running smoothly for dozens of days.

I notice that while I have checked to have Tugtainer check for updates, it seems that it is able to find an update for the socket and downloads it. It just doesn't update it. I believe, it is due to the "Protected" setting?

That is expected behavior - socket-proxy cannot be updated from tugtainer, since it's used to communicate with the docker socket. The protected label was made just for this purpose.

[ALERT] (1) : Current worker (11) exited with code 143 (Terminated)

There might be something wrong with your config or something else. As far as I know, 143 is a graceful stop. Something is stopping the container. But what exactly is unclear. Maybe an external script, lack of resources, restart of docker daemon, watchtower or something else? You can use docker events or some other logs to try to figure this out.

For now i don't know how to help you.