Kubelet readOnlyPort is an unauthenticated cluster inventory path
The kubelet has a read-only port that serves requests with no authentication or authorization. The kubelet configuration documents the setting as readOnlyPort, and the command-line reference says to set it to 0 to disable it. A cluster that leaves this port reachable gives an unauthenticated caller a path to kubelet status and pod metadata that should normally sit behind identity and access controls.
This is not the same as the kubelet's secured HTTPS endpoint. The secured port can use client certificates, bearer tokens, anonymous authentication policy, and authorization through the API server. The read-only port bypasses those checks by design. Its name can make it sound harmless. In practice, pod names, namespaces, container images, node information, and runtime details are useful reconnaissance material and can expose the shape of a production environment.
Why the setting survives
Kubelet configurations vary by distribution, installer, and cluster age. A tutorial may use the read-only port to simplify a local lab. A legacy configuration file may remain on a node after the managed platform changes its defaults. A security team may verify the API server's anonymous access settings and never inspect the node-level kubelet configuration.
The Kubernetes project's API server bypass guidance explicitly recommends ensuring that the unauthenticated read-only kubelet port is not enabled. That wording is useful because the risk is not limited to the API server's own port. A control-plane hardening review must include every node's kubelet configuration and the network path that reaches it.
Do not infer the setting from a provider's marketing page or from a single node. Inventory the effective kubelet configuration for each node pool and each image family. Check for command-line overrides as well as the configuration file. A configuration file that sets the value to zero can still be overridden by a stale command-line flag depending on how the node bootstrap composes its arguments.
Checks that do not collect cluster secrets
On a node you are authorized to inspect, check whether a listener exists on the read-only port. A local process listing is enough to identify the port in many deployments. Capture the owning process and the node identity, then compare the result with the intended configuration.
The Kubernetes API server bypass guidance also recommends checking whether anonymous requests to the secured kubelet endpoint are limited. Enable webhook authorization where the cluster design calls for it, keep anonymous authentication disabled or tightly constrained, and verify that the node's certificate and token paths reach the intended authenticator.
Network placement is the next control. Kubelet ports should not be reachable from the public internet or from broad user networks. Nodes should accept kubelet traffic only from the control plane, monitoring systems, and the specific administrative paths that require it. A firewall rule that allows an entire corporate range weakens the authentication boundary because it turns a compromised workstation into a kubelet client.
Do not scan the internet or test other organizations' nodes. Use an authorized inventory of your own addresses and a controlled test environment. Do not paste kubelet responses into a ticket or chat if they contain workload names, labels, or image references that are sensitive.
Remediation and verification
Set the read-only port to zero in the maintained kubelet configuration and remove any conflicting command-line override. Restart or reprovision through the supported platform method. Confirm that the listener is gone from every node, not only the node used for the first test. Check that the secured endpoint still has the intended authenticator and authorizer.
Review network policy and host firewall rules after the configuration change. A port that is no longer listening is no longer a data path, but the firewall should still deny unapproved access to the node API. Keep monitoring and control-plane traffic working, and verify that the node remains healthy after a rollout.
Add a configuration assertion to the node bootstrap or compliance pipeline. The assertion should fail when the read-only port is nonzero, when anonymous authentication is unexpectedly enabled, or when the node certificate and token authentication paths are missing. Treat a configuration drift alert as a security event, not only an availability warning.
Top comments (0)