Basic Kubernetes Commands

1. Running an Interactive Pod

To launch an ad-hoc debugging or exploration session with an interactive terminal:

kubectl run test-session \
  --namespace sci-myproject \
  --image=ubuntu:22.04 \
  --restart=Never \
  -it -- /bin/bash

---

2. Deploying a Manifest (.yaml)

Scientific workloads are defined using declarative YAML manifests.

Save the following as my-pod.yaml:

apiVersion: v1
kind: Pod
metadata:
  name: gpu-analysis-test
  namespace: sci-myproject
spec:
  restartPolicy: OnFailure
  containers:
    - name: worker
      image: registry.berkelium.lbl.gov/sci-als/synchrotron-model:v1
      command: ["python", "app/evaluate.py"]
      resources:
        limits:
          cpu: "8"
          memory: "32Gi"
          nvidia.com/gpu: "1"
        requests:
          cpu: "4"
          memory: "16Gi"
          nvidia.com/gpu: "1"

Apply and monitor the pod:

# Submit pod to the cluster
kubectl apply -f my-pod.yaml

# Check status
kubectl get pods -n sci-myproject

# Follow logs in real-time
kubectl logs -f gpu-analysis-test -n sci-myproject

---

3. Useful Cheat Sheet

# Get detailed status and scheduling events
kubectl describe pod <pod-name> -n sci-myproject

# Open a shell inside an active container
kubectl exec -it <pod-name> -n sci-myproject -- /bin/bash

# View resource usage (CPU & Memory)
kubectl top pods -n sci-myproject

# Delete a completed or obsolete pod
kubectl delete pod <pod-name> -n sci-myproject