fix(metrics): kube-state-metrics pod metadata

[swiatekm-sumo]

We always drop container and pod attributes from kube state metrics if their value is the kube-state-metrics pod/container itself. This is usually wrong, as the attributes are added by the ServiceMonitor configuration to metrics which aren't about pods to begin with, like kube_deployment_status_replicas, for example.

However, this means we incorrectly drop these attributes from pod metrics about the kube-state-metrics pod itself, like kube_pod_status_phase.

This PR fixes this bug by checking for the uid attribute, which kube-state-metrics adds to all metrics about pods.

Checklist

• Changelog updated or skip changelog label added
• Integration tests added or modified for major features

[sumo-drosiek]

Nice!