Support node with ip:port and ip, related issue: #100

@qmhu , hello

1. I would like to know why scheduler chart appVersion is 0.2.0 instead of 0.1.23 or 0.0.23, what does 0.2.0 mean? I think the appVersion should be the same as the version of the main project, what do you think?
2. the last commit did not modify the chartVersion, which caused the CI process to report an error, the last modification was not applied, I can try to fix it.

version:
crane 0.7.0
k8s:1.22.10
helm chart version:0.7.0

i install crane to kube-system namespace,the apiserver and webhook will get errors,i found wenhook certs only cn=crane-system,and api server helm template have static namespace。
apiVersion: apiregistration.k8s.io/v1 kind: APIService metadata: name: v1beta1.external.metrics.k8s.io labels: {{- include "crane.labels" . | nindent 4 }} spec: service: name: metric-adapter namespace: crane-system group: external.metrics.k8s.io version: v1beta1 insecureSkipTLSVerify: true groupPriorityMinimum: 100 versionPriority: 100

老师，您好:
如何使用自有Prometheus+Kafka搭建Crane？我修改了Grafana和Prometheus的连接地址，但是现在Crane控制台无法获取到Grafana。并且使用 [域名]:[port]/grafana无法正确跳转的我们Grafana页面。请知道。

By default, kube-state-metrics will override namespace label by itself deployed, the real workload namespace is exported_namespace, this is unexpected for us.

prometheus-community/helm-charts#1368

because generally, we use namespace to denote the workload namespace.

both usage and request is wrong.

nginx need configure Access-Control-Allow-Origin to allow one crane dashboard to access other clusters grafana and dashboard

Move agent tsp configmap to crane-system namespace, it should not be in the default ns