Connecting fluentd timeout causes openstack project startup problems

I encountered one such bug.I use k8s to deploy my application,and send the log to fluentd.However, when fluentd's pod dies, fluentd's service is still normal.At this time, my openstack application will not start up and will keep trying to connect fluentd's service.
Is there anyone who can help me?

No file, test, or entry point is identified in the issue. Reproduce the Kubernetes case where the Fluentd pod dies while its service remains available, then trace the Python logger's connection and timeout behavior; done means the OpenStack application no longer hangs indefinitely when Fluentd is unavailable.

由索引模型根据 Issue 内容生成。