阿里云配置AI组件报错
背景：手动配置过一次GPU共享插件，但失败了，结果残留了部分配置没修改，还有部分进程未关闭，导致一键配置失败，经过如下步骤后可以重新一键配置。![[Error: rendered manifests contain a resource that already exists. Unable to continue with install: existing resource conflict: kind: Service, namespace: default, name: gpushare-plugin]](http://jalon.link/usr/uploads/2022/07/45434294.png "[Error: rendered manifests contain a resource that already exists. Unable to continue with install: existing resource conflict: kind: Service, namespace: default, name: gpushare-plugin]")
1、删除SA(Service Account)
kubectl delete sa [sa_name] -n kube-system2、删除clusterrole
kubectl delete clusterrole [clusterrole_name]3、删除clusterrolebinding
kubectl delete clusterrolebinding [clusterrolebinding_name]4、删除deamonset
kubectl delete deamonset [deamonset_name] -n kube-system