I’ve been annoyed recently with a bunch of junk in my dmesg output:
[Thu Jul 11 10:29:27 2024] docker0: port 5(veth9509698) entered blocking state
[Thu Jul 11 10:29:27 2024] docker0: port 5(veth9509698) entered disabled state
[Thu Jul 11 10:29:27 2024] veth9509698: entered allmulticast mode
[Thu Jul 11 10:29:27 2024] veth9509698: entered promiscuous mode
[Thu Jul 11 10:29:27 2024] docker0: port 5(veth9509698) entered disabled state
[Thu Jul 11 10:29:27 2024] veth9509698 (unregistering): left allmulticast mode
[Thu Jul 11 10:29:27 2024] veth9509698 (unregistering): left promiscuous mode
[Thu Jul 11 10:29:27 2024] docker0: port 5(veth9509698) entered disabled state
This repeats every couple seconds ad infinitum and pollutes the log making it hard when it’s time to debug other things.
I finally decided to track it down and found a few different solutions but none of them worked. I finally found a mention in a bug report that it was normal. Surely it’s not normal to spew messages constantly, though… right?
After pondering it a bit and re-reading some context it seemed like they were saying it was normal to print that every time a container started, not just printing it constantly. That led me to thinking, what if I’ve got a container that’s constantly starting and stopping? What would cause that and how do I find it?
Poking around in the docker command I found docker events, which sounded promising. Sure enough, after running it I saw that a container was constantly starting and stopping. It turns out this container was a failed attempt at something and it had a systemd service that was constantly trying to get it to run. When I stopped working on it I forgot remove the .service file from /etc/systemd/system. Once I stopped the service and deleted the service file, the spewing stopped. Ahhhhhhh.