feat: monitor etcd

[uhthomas]

It looks like etcd does not run in-cluster with Talos, and therefore is not easy to monitor.

[uhthomas]

name: "vm-victoria-metrics-k8s-stack-kube-etcd"
spec: {
    selector: matchLabels: {
        app:                          "vm-victoria-metrics-k8s-stack-kube-etcd"
        "app.kubernetes.io/instance": "vm"
    }
    namespaceSelector: matchNames: [
        "kube-system",
    ]
    endpoints: [{
        bearerTokenFile: "/var/run/secrets/kubernetes.io/serviceaccount/token"
        port:            "http-metrics"
        scheme:          "https"
        tlsConfig: caFile: "/var/run/secrets/kubernetes.io/serviceaccount/ca.crt"
    }]
    jobLabel: "jobLabel"
}