Kubernetes CLI Plugin - Doctor
This plugin is inspired from brew doctor :) It will scan your currently target
ed k8s cluster to see if there are anomalies or useful action points that it can report back to you.
This plugin does not change any state or configuration, it merely just scans and gathers information than reports back anomalies in yaml format.
Install
- Download a zip that contains the binary from releases that is compatible with your os/arch
- Unzip to get
kubectl-doctor
(orkubectl-doctor.exe
if windows) - Add it to your
PATH
Usage
When the plugin binary is found from PATH
you can just execute it through kubectl
CLI
kubectl doctor
Current list of anomaly checks
- core component health (etcd cluster members, scheduler, controller-manager)
- orphan endpoints (endpoints with no ipv4 attached)
- persistent-volume available & unclaimed
- persistent-volume-claim in lost state
- k8s nodes that are not in ready state
- orphan replicasets (desired number of replicas are bigger than 0 but the available replicas are 0)
- leftover replicasets (desired number of replicas and the available # of replicas are 0)
- orphan deployments (desired number of replicas are bigger than 0 but the available replicas are 0)
- leftover deployments (desired number of replicas and the available # of replicas are 0)
- leftover cronjobs (last active date is more than 30 days)