[banyandb]Observable metric error

[Hwting]

HI, Observable metric error
version：0.11.1
deploy：helm
1.Prometheus metrics collected via the Fodc proxy are inaccurate.

My role: The data memory limits are as follows:

            limits:
              cpu: '4'
              memory: 20000Mi
            requests:
              cpu: '2'
              memory: 8000Mi

Skipping the FODC proxy collection role and validating metrics directly against the data node confirms the metrics are actually correct, indicating the issue lies with the FODC proxy.