📊 Monitor Pods
Monitoring a pod means answering three questions: is it running? Is it stable (not restarting)? Is it consuming a lot of resources?
info
Pod monitoring is done with kubectl. The Nuvion Panel shows the cluster and the nodes, but does not have a pods screen. For CPU, memory and nodes, see the cluster monitoring pages.
🚀 Before you start
- Have cluster access configured (see Cluster Access).
- Usage data (
kubectl top) relies on the Metrics Server, which is already installed in the cluster.
🚀 1. See the pod status
kubectl get pods -A
Look at two columns:
| Column | What it shows |
|---|---|
| READY | Ready containers / total (1/1 is ideal) |
| RESTARTS | How many times the container restarted |
To follow in real time, add --watch:
kubectl get pods --watch
Exit with Ctrl+C.
🚀 2. See CPU and memory usage
kubectl top pods -A
CPU appears in m (millicores: 250m is a quarter of a CPU) and memory in Mi.
To sort by the biggest consumers:
kubectl top pods -A --sort-by=memory
kubectl top pods -A --sort-by=cpu
🚀 3. Find pods with problems
kubectl get pods -A --field-selector=status.phase!=Running,status.phase!=Succeeded
The command lists only the pods that are not running normally.
🚀 4. Investigate a pod
kubectl describe pod POD-NAME
See the Status, the Last State (reason for the last restart) and the Events section at the end. For logs, see the Logs page.
🚀 5. See usage per container
kubectl top pod POD-NAME --containers
✅ How to know it is healthy
- STATUS
Runningand READY complete (1/1,2/2). - RESTARTS stable (does not increase with each query).
- Usage below the limits defined in the Deployment.
❗ Common problems
| Symptom | Cause | Solution |
|---|---|---|
error: Metrics API not available | Metrics Server has not collected data yet | Wait 1 to 2 minutes and try again |
| RESTARTS increasing | The container fails or runs out of memory | See describe (Last State field, reason OOMKilled means out of memory) and the Logs |
Pending | No resources on the nodes | See Events and capacity in Scale |
CrashLoopBackOff | The container starts and fails repeatedly | See kubectl logs POD-NAME --previous |
➡️ Next step: Logs
🧠 Questions?
Talk to our SaveinCloud technical support team!