故障排除 OpenShift Operator (OLM) (非集群默认的 Operator)

Operator Lifecycle Manager

Operator Lifecycle Manager 缩写 OLM

官网上的工作流

OLM 负责管理非默认 operator 即从 Operator Hub 里安装来的Operator

查看OLM 状态.

oc get clusteroperator

如果这三个的状态有误, 就有可能是subs, csv, installplan出现问题或者错误配置OLM

命名空间 openshift-operator-lifecycle-manager

oc get all -n openshift-operator-lifecycle-manager

查看这两个 pod 详情

oc describe pod/catalog-operator-xxxxxxxxx-xxxxx -n openshift-operator-lifecycle-manager

oc describe pod/olm-operator-xxxxxxxxx-xxxxxx -n openshift-operator-lifecycle-manager

查看 Conditions

查看logs, 去看哪里失败了

oc logs catalog-operator-xxxxxxxx-xxxxx -n openshift-operator-lifecycle-manager

oc logs olm-operator-xxxxxxxx-xxxxx -n openshift-operator-lifecycle-manager

Cluster Service Version (csv) 集群服务版本

csv 是一个 yaml 文件, 用来协助 OLM去运行集群中的 Operator
也用于在用户界面添加信息,比如说版本,描述,logo

oc get csv

查看是所有的
查看 csv

也可以查看这个operator 的 pod , 以此来找到命名空间

oc get pods -A | grep <csv name>

oc get csv -o yaml

Subscription 查看订阅状态

oc get subs -n <operator namespace>

以我的为例

oc describe sub < package name > -n < operator namespace >

查看 Condition , 确实所有的CatalogSource 健康
状态是 false 说明所有

再或者, 查看 sub

oc get subs -n <operator namespace>

oc get subs <subname> -o yaml -n < operator namespace>

installplan

查看 operator 是手动更新还是自动更新

oc get installplan -A

installplan 也是表现着订阅状态的

Subs 里面也包含了 installplan 信息

oc get subs <subname> -o yaml -n < operator namespace>

或者查看 InstallPlan 的状态中的错误

apiVersion: operators.coreos.com/v1
kind: InstallPlan
status:conditions:- lastTransitionTime: "2019-07-26T21:13:10Z"lastUpdateTime: "2019-07-26T21:13:10Z"message: 'error creating clusterrole etcdoperator.v0.9.4-clusterwide-dsfx4: clusterroles.rbac.authorization.k8s.iois forbidden: User "system:serviceaccount:scoped:scoped" cannot create resource"clusterroles" in API group "rbac.authorization.k8s.io" at the cluster scope'reason: InstallComponentFailedstatus: "False"type: Installedphase: Failed

如果是因为缺少权限而导致 operator 安装失败, 可能会出现以上错误