helm部署的时候，grafana的initContainer出现报错日志: can't create directory '/var/lib/grafana/plugins/deepflow-apptracing-panel': No such file or directory...

[liuxuzxx]

1. 通过helm安装deepflow组件
2. 其他组件:clickhouse,mysql,app,server,agent都正常，但是grafana的initContainer出现错误日志如下:

cp: can't create directory '/var/lib/grafana/plugins/deepflow-apptracing-panel': No such file or directory
cp: can't create directory '/var/lib/grafana/plugins/deepflow-querier-datasource': No such file or directory
cp: can't create directory '/var/lib/grafana/plugins/deepflow-topo-panel': No such file or directory
cp: can't create directory '/var/lib/grafana/plugins/grafana-clickhouse-datasource': No such file or directory

目前我的临时解决方案是:

1. 修改如下配置

---
apiVersion: apps/v1
...省略
spec:
  template:
    spec:
      initContainers:
        - image:  xwharbor.wxchina.com/cpaas/deepflow-ce/deepflowio-init-grafana:v7.0.4
        
          #1.增加挂载，这个点有点疑惑啊，pod内的container不是通过volumeMounts来共享文件系统的吗？之前都没有volumeMounts，如何共享的?
          volumeMounts:
            - mountPath: /var/lib/grafana/plugins
              name: custom-plugins
      #2.直接粗暴的给root级别的用户运行，先跑起来在说
      securityContext:
        runAsUser: 0

[liuxuzxx]

仔细看了下，是在依赖的charts，这个grafana里面的_pod.tpl这个定义的，需要一些values.yaml设置，才能执行如下的几个init-containers:

1. init-chown-data(应该是可以直接使用472这个用户级别创建文件夹)
2. download-dashboards这个看着是下载一些dashboards
3. xxx-init-sc-alerts 应该是和告警相关
4. xxx-init-sc-datasources 可能是配置数据源初始化的动作
5. xxx-init-sc-notifiers 看着像是通知