Feature Request: Add metrics related to kubernetes_logs rotate_wait_secs

[davidcpell]

Discussed in #22060

^{Originally posted by davidcpell December 19, 2024}
I've recently turned on rotate_wait_secs in my Kubernetes environments because I was running into issues where Vector would be holding tons of file handles for files that have passed rotation size but hasn't finished processing them due to the amount of work it's doing. This led to node DiskPressure over time and pod evictions.

Using this helped, but it'd be really helpful to be able to see how often Vector is dropping file handles early and any possible metadata related to which files (container name, whatever). Is this available/something that could be reasonably added?