Braden Schaeffer

> I'm not sure how we would fix this without completely rearchitecting things to not rely on the kubernetes logging system. We don't have any plans to do this right...

From #11511: > FWIW, this does seem to be related to container runtime. In GKE, we recently upgraded to 1.20 which defaults to containerd at the same time we saw...

So, we're just starting a project and were looking to use swarm with this exact use case in mind. Is this currently still an issue?

Hi everyone... is there a workaround for this?

@mrsimo Yes. With load balancing alone and also within the last 24h

> Apparently this can be fixed by specifying: --no-collector.fill-missing-labels. Not sure about any other implications. This did not work for us

@AlanCoding It is possible with the API though possibly https://docs.ansible.com/ansible/latest/collections/awx/awx/schedule_module.html

Hrm... still seems like an issue

Just want to say that increasing size before log rotation does not solve the problem. It just decreases the likelyhood that your run will take place or trigger a log...

+1