The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Kubernetes node heartbeats and container readiness probes answer different questions. A node heartbeat helps the control plane determine whether a node is available; a readiness probe tells Kubernetes whether a particular container is ready to receive traffic. A failed readiness probe removes the Pod from matching Service endpoints but does not restart it. Liveness probes, not readiness probes, can trigger a container restart.
What Kubernetes node health checks measure
Kubernetes uses node-level signals to reason about whether a machine is available to the cluster. Its documentation describes two heartbeat forms: updates to a Node’s .status and a Lease associated with that Node in the kube-node-lease namespace. The kubelet reports these signals; the control plane uses node availability information to detect failures and respond. Kubernetes describes heartbeats as helping the cluster determine node availability and take action when failures are detected.
As an Amazon Associate I earn from qualifying purchases.
The Node Ready condition is the node-level health signal relevant to accepting Pods. It does not report whether an individual application container can serve requests. That application-level question is handled by container probes.
Recommended Free Tools
How Node status heartbeats and node Leases differ
| Signal | Scope and meaning | How it is updated | Effect |
|---|---|---|---|
| Node status heartbeat | Node-level status and conditions, including Ready. |
The kubelet posts Node status when it changes or at an interval. | The node controller uses node availability information to detect failures and take action. |
| Node Lease heartbeat | A lightweight indication of liveness for the associated Node. | The kubelet creates and renews a Lease in kube-node-lease, independently of Node status updates. |
It helps the cluster assess node availability with less update impact in large clusters. |
A Lease is not a replacement for Node status: the two signals serve related but distinct roles. Kubernetes’ v1.35 Node Status reference calls a Lease lightweight compared with Node .status updates and explains that Lease updates happen independently.
What happens when a node stops reporting
If the control plane does not hear from a node within the configured node-monitor grace period, the Node’s Ready condition can become Unknown. The Kubernetes v1.35 Node Status reference lists 50 seconds as the default for node-monitor-grace-period; the effective value can differ with release or cluster configuration. This is a node-level judgment, not the result of a readiness probe.
Loss of heartbeats is used to detect node availability problems, but it does not imply one universal time for Pod eviction. Subsequent behavior depends on controller timing, node conditions and taints, Pod tolerations, and configuration. Check the documentation and effective settings for the Kubernetes release or managed service in use before relying on a specific recovery timeline.
What readiness probes do to a Pod and its traffic
A readiness probe checks whether a container is ready to accept traffic. The kubelet periodically runs the configured check. When readiness fails, the container continues running, but the Pod becomes unready; its IP is removed from EndpointSlices for matching Services. The check therefore changes whether the workload is eligible for Service traffic, rather than declaring the node unavailable.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesKubernetes supports HTTP, TCP, exec, and gRPC probe mechanisms. The endpoint or test should reflect the application’s actual ability to serve requests. A healthy node can host a container that fails readiness, just as a node-level heartbeat problem can occur independently of a container’s readiness result.
Readiness, liveness, and startup probes compared
| Probe or signal | Question it answers | Failure consequence |
|---|---|---|
Node heartbeat / Node Ready |
Is the node available and able to accept Pods? | Control-plane node lifecycle behavior follows node condition and cluster configuration. |
| Readiness probe | Can this container accept traffic now? | The Pod becomes unready and is removed from matching Service EndpointSlices; the container is not restarted. |
| Liveness probe | Should this container be considered unhealthy and restarted? | When failures reach the configured threshold, the kubelet can restart the container. |
| Startup probe | Has a slow-starting application initialized enough for regular health checks? | It can gate liveness and readiness checks until startup succeeds. |
Readiness and liveness checks may use similar mechanisms, but their failure actions are different. Choose readiness criteria that represent the application’s ability to handle traffic. Use liveness for a condition that indicates the container should be restarted, rather than for temporary load or dependency problems that may resolve without a restart. Kubernetes warns that incorrectly implemented liveness probes can contribute to cascading failures: restarts under load can add demand to the remaining Pods.
Timing defaults: interpret them in their documented context
These are reference values, not guarantees for every cluster. Verify the Kubernetes release and effective kubelet and controller configuration before using them to calculate an operational deadline.
| Setting or behavior | Documented value | Context |
|---|---|---|
| Lease update interval | 10 seconds | Default listed in the Kubernetes v1.35 Node Status reference. |
| Lease update retry backoff | Starts at 200 milliseconds; capped at 7 seconds | Retry behavior described in the Kubernetes v1.35 Node Status reference. |
| Node status updates when status is unchanged | 5 minutes | Default interval described in the Kubernetes v1.35 Node Status reference. |
--node-status-update-frequency |
10 seconds | Default in the kubelet command reference; the reference cautions that it must work with the node controller’s nodeMonitorGracePeriod. |
node-monitor-grace-period |
50 seconds | Default listed in the Kubernetes v1.35 Node Status reference; check the cluster’s actual value. |
The 10-second command-line flag default and the five-minute unchanged-status interval describe different documentation contexts; do not combine them into a universal Node status update cadence.
For probe configuration, Kubernetes documents these defaults: periodSeconds is 10 seconds, timeoutSeconds is 1 second, successThreshold is 1, and failureThreshold is 3. These fields configure checks and thresholds. The threshold alone does not establish an exact end-to-end time to a traffic change or restart: scheduling, probe execution, and configured termination grace also affect when action occurs. See the probe configuration reference.
How to choose the signal to investigate
- Several workloads on one node are affected: inspect node conditions and heartbeat-related status, including the Node’s
Readycondition and Lease, as well as the cluster’s configured monitoring and recovery behavior. - One Pod is not receiving Service traffic while its node is healthy: inspect the Pod’s readiness probe results and its Service EndpointSlices.
- A container repeatedly restarts: inspect liveness probe configuration and events; readiness failure by itself does not restart the container.
- A workload starts slowly: consider whether a startup probe should gate liveness and readiness checks until initialization completes.
For production tuning, use the exact Kubernetes release and provider documentation: Lease renewal, Node status updates, grace periods, taints, tolerations, and eventual Pod eviction can vary with configuration.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
