Testing API from K3s on Pi5 Master

Your Threat Detection System is running on K3s with 9 API replicas distributed across the cluster!

Current K3s Deployment Status

PODS RUNNING:
✅ tds-api:      9 replicas (across cluster nodes)
✅ tds-postgres: 1 replica (pi5-master)
✅ tds-minio:    4 replicas (distributed)
✅ tds-mqtt:     1 replica (pi5-master)

---

Quick Test from Pi5

Run Auto Test Script

ssh cc123@192.168.1.50
chmod +x Threat-Detection-System/test-api-from-k3s.sh
./Threat-Detection-System/test-api-from-k3s.sh

This script checks: - ✅ K3s cluster status - ✅ Pod health - ✅ Service discovery - ✅ API connectivity - ✅ Detection events - ✅ Statistics - ✅ MQTT broker - ✅ Database status

---

Manual Testing from Pi5

1. SSH to Pi5

ssh cc123@192.168.1.50

2. Check Deployment Status

# All TDS pods
kubectl get pods -l 'app in (tds-api, tds-postgres, tds-minio, tds-mqtt)'

# Detailed info
kubectl get pods -l 'app in (tds-api, tds-postgres, tds-minio, tds-mqtt)' -o wide

3. Test API (Internal Access)

Via Service DNS:

# Health check
curl http://tds-api:8001/

# Get recent detections
curl http://tds-api:8001/api/v1/detections/recent?limit=5 | jq

# Get statistics
curl http://tds-api:8001/api/v1/detections/statistics | jq

# Filter by severity
curl http://tds-api:8001/api/v1/detections?severity=critical | jq

Via ClusterIP:

# Get ClusterIP
kubectl get svc tds-api

# Use the IP
curl http://<CLUSTERIP>:8001/api/v1/detections

4. Check Service Details

# API service
kubectl get svc tds-api -o wide

# All services
kubectl get svc -l 'app in (tds-api, tds-postgres, tds-minio, tds-mqtt)'

---

Access from Your Workstation

Option 1: Port-Forward from Pi5

ssh cc123@192.168.1.50

# In one terminal - keep it running
kubectl port-forward svc/tds-api 8001:8001 &

# Then on your workstation access:
# http://localhost:8001/docs

Option 2: NodePort Access

ssh cc123@192.168.1.50

# Get the NodePort
kubectl get svc tds-api -o jsonpath='{.spec.ports[0].nodePort}'

# Then from your workstation:
curl http://192.168.1.50:NODEPORT/api/v1/detections

Option 3: LoadBalancer (If Configured)

ssh cc123@192.168.1.50
kubectl get svc tds-api

# If EXTERNAL-IP is assigned:
curl http://EXTERNAL_IP:8001/api/v1/detections

---

Complete Testing Workflow

Terminal 1: Monitor Logs

ssh cc123@192.168.1.50
kubectl logs -f -l app=tds-api

Terminal 2: Monitor MQTT Messages

ssh cc123@192.168.1.50
mosquitto_sub -h localhost -t 'cluster/camera/events' -v

Terminal 3: Send Camera Stream from Pi4/Pi5

# On Pi4 edge node or another machine
python3 edge_camera_publisher.py \
  --broker 192.168.1.50 \
  --enable-detections \
  --fps 5 \
  --sensor-id "pi4-edge-001"

Terminal 4: Monitor Detection Events

ssh cc123@192.168.1.50

# Watch events in real-time
while true; do
  echo "=== $(date) ==="
  curl -s http://tds-api:8001/api/v1/detections/recent?limit=3 | jq '.[] | {event_type, severity, confidence}'
  sleep 5
done

---

Swagger UI Access

From Workstation via Port-Forward

# Terminal 1: Create port-forward
ssh cc123@192.168.1.50
kubectl port-forward svc/tds-api 8001:8001 &

# Terminal 2: Open in browser
# http://localhost:8001/docs

From Pi5 Terminal (Check for issues)

ssh cc123@192.168.1.50
curl http://tds-api:8001/docs

---

Database Access

Query Detection Events

ssh cc123@192.168.1.50

# Port-forward PostgreSQL
kubectl port-forward svc/tds-postgres 5432:5432 &

# On your workstation, connect:
psql -h localhost -U tds_user -d threat_detection
# Password: tds_secret

# SQL queries
SELECT id, event_type, severity, confidence, detected_at 
FROM detection_events 
ORDER BY created_at DESC 
LIMIT 10;

# Check images
SELECT de.id as event_id, di.storage_key, di.file_size_bytes
FROM detection_events de
JOIN detection_images di ON di.detection_event_id = de.id
ORDER BY de.created_at DESC;

---

MQTT Broker Access

Monitor Messages from Pi5

ssh cc123@192.168.1.50

# Subscribe to camera events
mosquitto_sub -h localhost -t 'cluster/camera/events' -v

# Subscribe to all topics
mosquitto_sub -h localhost -t '#' -v

Publish Test Message

ssh cc123@192.168.1.50

mosquitto_pub -h localhost -t 'cluster/camera/events' -m '{"node":"test","label":"person"}'

---

MinIO Image Storage

Access MinIO Dashboard

ssh cc123@192.168.1.50

# Get MinIO service info
kubectl get svc -l app=tds-minio

# Port-forward MinIO console
kubectl port-forward svc/tds-minio-console 9001:9001 &

# On your workstation:
# http://localhost:9001
# Login: admin / password123

---

Troubleshooting

API Pod Not Running

ssh cc123@192.168.1.50

# Check pod status
kubectl get pods -l app=tds-api -o wide

# View logs
kubectl logs -l app=tds-api --tail=50

# Describe pod (shows events)
kubectl describe pod <pod-name>

API Slow or Unresponsive

ssh cc123@192.168.1.50

# Check resources
kubectl top pods -l app=tds-api

# Check node resources
kubectl top nodes

# Restart deployment
kubectl rollout restart deployment/tds-api

Database Connection Issues

ssh cc123@192.168.1.50

# Check database logs
kubectl logs -l app=tds-postgres --tail=50

# Port-forward and test connection
kubectl port-forward svc/tds-postgres 5432:5432 &
psql -h localhost -U tds_user -d threat_detection

MQTT Not Receiving Messages

ssh cc123@192.168.1.50

# Check MQTT logs
kubectl logs -l app=tds-mqtt --tail=50

# Test MQTT connectivity
mosquitto_sub -h localhost -t 'test/topic' &
mosquitto_pub -h localhost -t 'test/topic' -m 'test'

---

Performance Monitoring

View Real-Time Metrics

ssh cc123@192.168.1.50

# CPU and Memory usage
watch kubectl top pods -l 'app in (tds-api, tds-postgres, tds-minio, tds-mqtt)'

# Node usage
watch kubectl top nodes

Check API Metrics

# If Prometheus is installed
curl http://tds-api:8001/api/v1/metrics

---

Common Commands Reference

Command	Purpose
kubectl get pods	List all pods
kubectl get svc	List all services
kubectl logs -f <pod>	Stream logs
kubectl exec -it <pod> bash	Shell into pod
kubectl port-forward <service> <port>	Forward port
kubectl rollout restart deployment/<app>	Restart deployment
kubectl top pods	Resource usage

---

Edge Publisher Configuration

Run from Pi4 or your machine with camera access:

python3 edge_camera_publisher.py \
  --broker 192.168.1.50 \
  --port 1883 \
  --enable-detections \
  --yolo-model yolov8n.pt \
  --confidence 0.5 \
  --fps 10 \
  --sensor-id "pi4-edge-camera"

---

What's Running Where

Component	Pod	Node	Service IP	Port
API (x9)	tds-api-*	workers + master	10.42.x.x	8001
PostgreSQL	tds-postgres-*	pi5-master	10.42.0.x	5432
MQTT	tds-mqtt-*	pi5-master	10.42.0.x	1883
MinIO (x4)	tds-minio-*	distributed	10.42.x.x	9000

---

Quick Start

# 1. SSH to Pi5
ssh cc123@192.168.1.50

# 2. Run test script
./Threat-Detection-System/test-api-from-k3s.sh

# 3. Port-forward API (if you want local access)
kubectl port-forward svc/tds-api 8001:8001 &

# 4. Access Swagger UI
# Local: http://localhost:8001/docs
# Or: http://192.168.1.50:NODEPORT/docs (after checking NodePort)