clustercheck is a little helper to check the current health of a Kubernetes cluster.
It checks pre-defined KPI while query Prometheus backend. For that some requirements are needed:
- access to Prometheus from command line, between https_proxy variable
- configured
kubeconfig, clustercheck will use the current context
We provide binaries for various platform. Go to the release page.
% clustercheck
APISERVER 🟢 OK (1)
CLUSTER 🟢 OK (1)
FLUENTBITERRORS 🔴 FAIL (0)
FLUENTDERRORS 🟢 OK (1)
GOLDPINGER 🔴 FAIL (0)
KUBEDNS 🟢 OK (1)
KUBELET 🟢 OK (1)
NETWORKOPERATOR 🟢 OK (1)
NODE 🟢 OK (1)
STORAGECHECK 🟢 OK (1)
PROMETHEUSAGENT 🔴 FAIL (0)
SYSTEMPODS 🟢 OK (1)
xattr -d com.apple.quarantine $HOME/bin/clustercheck
export CLUSTER="my-cluster"
export PROMETHEUS_URL="https://my-prometheus.instance"
export PROM_USER="user"
export PROM_PASS="pass"
Start the programm with -bw or set env var
export CLUSTERCHECK_BW=1
In this version the programm expect an item on a Bitwarden service containing username/password for HTTP Basic Auth on Prometheus API
bw get item "Prometheus Agent RemoteWrite
If your cluster has a FQDN which is specific to set start the programm with -f together with the FQDN or set env var
CLUSTERCHECK_FQDN=example.com
we skip SSL verification and allow insecure connection by default, take care.
we respect env vars like http_proxy or https_proxy for Prometheus endpoint connection from your computer.