Found some potential issue via kubeeye

kubeeye is a tool that can help us to find some potential issues of the Kubernetes components. I use it to scan one of my clusters, got the following output:

root@node1:~# ke diag | grep devops
I0825 13:43:10.696334   24926 request.go:621] Throttling request took 1.006868293s, request: GET:https://lb.kubesphere.local:6443/apis/snapshot.storage.k8s.io/v1beta1?timeout=32s
ERRO[0001] Cache missed NotificationManager/kubesphere-monitoring-system/notification-manager again
ERRO[0001] Cache missed ClusterRoleBinding/kubesphere-controls-system/admin-cluster-admin again
default       Warning    admin.169e6d51b924ffb3                         2021-08-25T13:37:58+08:00   FailedSync   Failed to sync: Get "http://ks-jenkins.kubesphere-devops-system.svc/role-strategy/strategy/getRole/?roleName=admin&type=globalRoles": dial tcp: lookup ks-jenkins.kubesphere-devops-system.svc on 169.254.25.10:53: no such host
kubesphere-devops-system       Warning    migrate                           Job                   2021-08-25T13:43:11+08:00   [memoryRequestsMissing runAsRootAllowed cpuLimitsMissing livenessProbeMissing readinessProbeMissing cpuRequestsMissing pullPolicyNotAlways notReadOnlyRootFilesystem memoryLimitsMissing priorityClassNotSet]
kubesphere-devops-system       Warning    devops-ks-devops-apiserver        Deployment            2021-08-25T13:43:11+08:00   [cpuRequestsMissing readinessProbeMissing cpuLimitsMissing livenessProbeMissing memoryRequestsMissing runAsRootAllowed memoryLimitsMissing notReadOnlyRootFilesystem priorityClassNotSet]
kubesphere-devops-system       Warning    devops-ks-devops-controller       Deployment            2021-08-25T13:43:11+08:00   [memoryRequestsMissing readinessProbeMissing cpuLimitsMissing memoryLimitsMissing notReadOnlyRootFilesystem runAsRootAllowed livenessProbeMissing cpuRequestsMissing priorityClassNotSet]
kubesphere-devops-system       Warning    devops-jenkins                    Deployment            2021-08-25T13:43:11+08:00   [notReadOnlyRootFilesystem runAsRootAllowed priorityClassNotSet]
kubesphere-devops-system       Warning    s2ioperator                       StatefulSet           2021-08-25T13:43:11+08:00   [pullPolicyNotAlways notReadOnlyRootFilesystem livenessProbeMissing readinessProbeMissing runAsRootAllowed priorityClassNotSet]

Luckily, I didn't find any higher severity problems. But it is worth it to scan it regularly. For example, we can scan it with every single PR.

The issue names kubeeye's ke diag command and proposes scanning every pull request. Start by checking how that command could run in this repository's pull-request workflow, then define which findings should be reported or block a PR. Done means the scan runs regularly and its expected behavior is documented.