Skip to content
Latchkey

cilium install: Deploy Cilium From CI

cilium install deploys Cilium into the current-context cluster using the cilium-cli, choosing sane defaults for the detected environment.

In ephemeral CI clusters (kind, k3d, minikube), cilium install is the quickest way to stand up the CNI before running connectivity tests.

What it does

cilium install detects the Kubernetes distribution and installs Cilium with an appropriate configuration, applying the agent DaemonSet and operator. You can pin the version and override Helm values with --set. It targets the cluster in your current kube context.

Common usage

Terminal
cilium install --version 1.15.6
# enable Hubble at install time
cilium install --set hubble.relay.enabled=true --set hubble.ui.enabled=true
# wait for readiness
cilium install --version 1.15.6 && cilium status --wait

Options

FlagWhat it does
--version <ver>Pin the Cilium version to install
--set <path=value>Override a Helm value (repeatable)
--values <file>Supply a Helm values file
--namespace <ns>Install namespace (default kube-system)
--dry-runShow what would be installed without applying

In CI

Pin --version so a floating latest does not change CNI behavior between runs. Follow install with cilium status --wait so the pipeline blocks until Cilium is ready before it schedules workloads or runs connectivity tests.

Common errors in CI

"Error: unable to install Cilium: ... context ... not found" means KUBECONFIG is wrong or empty. "detected ... but no matching ... configuration" can appear on unusual distros; pass explicit --set values. If a second install runs, "Cilium is already installed" tells you to use cilium upgrade instead.

Related guides

Run this faster and cheaper on Latchkey managed runners. Start free →