fix(metrics): drop cadvisor container metrics without metadata

[swiatekm-sumo]

Cadvisor container metrics are emitted for all cgroups, even ones not actually related to containers run by the kubelet. For some of these, the only label is the id containing the cgroup path:

$ kubectl get --raw "/api/v1/nodes/sumologic-control-plane/proxy/metrics/cadvisor" | grep container_cpu_usage_seconds_total
# TYPE container_cpu_usage_seconds_total counter
container_cpu_usage_seconds_total{container="",cpu="total",id="/",image="",name="",namespace="",pod=""} 676.209622261 1678268968878
container_cpu_usage_seconds_total{container="",cpu="total",id="/docker/bc61f8197b2e99d91a3e3da589e9c13dbfd2756a4e7f9903a2b76c4e6b0b5fda/kubelet.slice/kubelet-kubepods.slice/kubelet-kubepods-besteffort.slice/kubelet-kubepods-besteffort-pod3d6b3bf5_5b1c_407c_8842_04565a00d150.slice/cri-containerd-3ac92040b58042a3828de42432948019484832e0f20e03da5c1b09b5bc7d8f13.scope",image="k8s.gcr.io/pause:3.6",name="3ac92040b58042a3828de42432948019484832e0f20e03da5c1b09b5bc7d8f13",namespace="kube-system",pod="kube-proxy-2ccmm"} 0.01729504 1678268965373
container_cpu_usage_seconds_total{container="",cpu="total",id="/docker/bc61f8197b2e99d91a3e3da589e9c13dbfd2756a4e7f9903a2b76c4e6b0b5fda/kubelet.slice/kubelet-kubepods.slice/kubelet-kubepods-besteffort.slice/kubelet-kubepods-besteffort-pod9b9253b0_57a3_4ead_97c4_fb77f5c3c9a6.slice/cri-containerd-1c4169329d281834a35068176f49a6491b77d54bd0284670a8af040ce03883f0.scope",image="k8s.gcr.io/pause:3.6",name="1c4169329d281834a35068176f49a6491b77d54bd0284670a8af040ce03883f0",namespace="local-path-storage",pod="local-path-provisioner-9cd9bd544-79qtd"} 0.049625352 1678268959599
container_cpu_usage_seconds_total{container="",cpu="total",id="/docker/bc61f8197b2e99d91a3e3da589e9c13dbfd2756a4e7f9903a2b76c4e6b0b5fda/kubelet.slice/kubelet-kubepods.slice/kubelet-kubepods-burstable.slice/kubelet-kubepods-burstable-pod4f661fbe2c2911480854bb5e7af7e447.slice/cri-containerd-6c5627a36b6a91e1c8539e128ae6a878ddb9fc74b5775a2dc67df028fb55bcf1.scope",image="k8s.gcr.io/pause:3.6",name="6c5627a36b6a91e1c8539e128ae6a878ddb9fc74b5775a2dc67df028fb55bcf1",namespace="kube-system",pod="etcd-sumologic-control-plane"} 0.056957712 1678268962995
container_cpu_usage_seconds_total{container="",cpu="total",id="/docker/bc61f8197b2e99d91a3e3da589e9c13dbfd2756a4e7f9903a2b76c4e6b0b5fda/kubelet.slice/kubelet-kubepods.slice/kubelet-kubepods-burstable.slice/kubelet-kubepods-burstable-pod5536bf0d7cbf6795298304a1be19d4e4.slice/cri-containerd-055241d4e9b99c69abdd34c9bedf2619900409d22b3a00835880331fb50a93f6.scope",image="k8s.gcr.io/pause:3.6",name="055241d4e9b99c69abdd34c9bedf2619900409d22b3a00835880331fb50a93f6",namespace="kube-system",pod="kube-apiserver-sumologic-control-plane"} 0.058457856 1678268971417
container_cpu_usage_seconds_total{container="",cpu="total",id="/docker/bc61f8197b2e99d91a3e3da589e9c13dbfd2756a4e7f9903a2b76c4e6b0b5fda/kubelet.slice/kubelet-kubepods.slice/kubelet-kubepods-burstable.slice/kubelet-kubepods-burstable-pod70ba3f84_8f81_43fa_80e6_65ceb572df07.slice/cri-containerd-87f5879164f9484fabbbd1a21e69cfaf16ee889af78d9a3e3fcf482991e4150d.scope",image="k8s.gcr.io/pause:3.6",name="87f5879164f9484fabbbd1a21e69cfaf16ee889af78d9a3e3fcf482991e4150d",namespace="kube-system",pod="coredns-6d4b75cb6d-j2q9k"} 0.052671836 1678268961640
...

We've used this setting internally for a while as well.

Checklist

• Changelog updated or skip changelog label added