Fluent bit metrics by pod/container

[adecchi-2inno]

I am using fluent chart version 0.30.3, and it is deployed as daemonset and a custom Prometheus installation.

At the moment, I am not able to get metrics based on pod name or container.

For example:
sum(rate(fluentbit_input_records_total{namespace="logging"}[1d])) by (pod)
sum(rate(fluentbit_input_records_total{namespace="logging"}[1d])) by (container)

If I switched pod by instance or node, I get the result, but my intention is to get by pod/container.
The idea is to try to understand the Pods with excessive Logs enabled, to try to understand the Pods that trigger my Index at Opensearch to increase heavily.

Is anything required to do at my custom Prometheus deployment ?