Add option to enable and expose Prometheus metrics internally to other pods

[patilsuraj767]

Only internally because usually Prometheus/Thanos are run inside the same Kubernetes cluster. So we need to just expose metrics from ipfs-cluster to other namespaces.