How to fix pods in Not Ready state?

kubectl get pods might show that some of the pods have 0/N value in READY column.

What is the meaning of READY column value?

In the context of kubectl get pods, the READY column shows the number of containers in the pod that have passed their health checks and are ready to serve traffic.

The anatomy of R/T:

• R (Left side): This is the number of containers currently Ready. A 0 means the application inside the container is not responding to its "Readiness Probe" or has not finished starting up.
• / (Separator): Separates ready containers from the total.
• T (Right side): This is the Total number of user containers defined in that pod.

When we have a pod in 0/1 state with a status of Running, it means the container has started, but Kubernetes does not consider it "healthy" enough to handle requests. Our pods are technically "alive" (Running) but "unusable" (Not Ready).

Common Reasons for 0/1 Running:

• Failed Readiness Probe: The application is running, but the health check URL (e.g., /ready) is returning an error or timing out.
• Slow Startup: The application takes a long time to initialize, and the "Initial Delay" isn't long enough.
• Dependency Issues: The pod is waiting for a database, a config file, or another service that isn't available.
• CrashLoopBackOff (Transitions): Sometimes pods flicker between 0/1 Running and 0/1 CrashLoopBackOff as they try to start and immediately fail.

How to find out exactly what's wrong

To diagnose why our my-app pods are stuck, run these two commands:

(1) Check the Events (Why it's not ready):

kubectl describe pod my-app-0 -n my-app-namespace

Look at the "Events" section at the bottom for messages like "Readiness probe failed."

Output example:

Events:

  Type     Reason     Age                   From     Message

  ----     ------     ----                  ----     -------

  Warning  Unhealthy  47s (x81205 over 8d)  kubelet  Readiness probe failed: HTTP probe failed with statuscode: 503

(2) Check the Logs (What the app is saying):

kubectl logs my-app-0 -n my-app-namespace

Look for "Error," "Exception," or "Connection Refused" messages.

---