Skip to content

Step 5: Visualization with Grafana

Learn how to create beautiful dashboards to understand your app

📊 What is Grafana?

Grafana is a visualization platform that creates dashboards from your metrics and logs:

Feature What It Does
Dashboards Visual panels showing metrics
Explore Ad-hoc queries for debugging
Alerting Visual alert rules
Datasources Connect to Prometheus, Loki

🌐 Accessing Grafana

  1. Open http://localhost:3000
  2. Login with:
  3. Username: admin
  4. Password: admin

🔍 Using Explore

Explore is great for ad-hoc queries:

Query Prometheus (Metrics)

  1. Click "Explore" in sidebar
  2. Select "Prometheus" datasource
  3. Enter a PromQL query:
# Request rate
rate(django_http_requests_total[5m])

# Latency
histogram_quantile(0.95, rate(django_http_requests_latency_seconds_bucket[5m]))

# CPU usage
100 - (avg(rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)
  1. Click "Run Query"

Query Loki (Logs)

  1. Click "Explore" in sidebar
  2. Select "Loki" datasource
  3. Enter a LogQL query:
# All Django logs
{app="django"}

# Error logs
{app="django", level="ERROR"}

# Search logs
{app="django"} |= "POST"
  1. Click "Run Query"

📈 Creating Your First Dashboard

Step 1: Create Dashboard

  1. Click "Dashboards" in sidebar
  2. Click "New Dashboard"
  3. Click "Add visualization"
  4. Select "Prometheus" datasource

Step 2: Add Request Rate Panel

  1. Enter query:
sum(rate(django_http_requests_total[5m]))
  1. Click "Run Query"
  2. In panel options:
  3. Title: "Request Rate"
  4. Unit: "requests/sec"
  5. Click "Apply"

Step 3: Add Latency Panel

  1. Click "Add panel"
  2. Enter query:
histogram_quantile(0.95,
  sum(rate(django_http_requests_latency_seconds_bucket[5m])) by (le)
)
  1. Set title: "P95 Latency"
  2. Set unit: "seconds"
  3. Click "Apply"

Step 4: Add Error Rate Panel

  1. Click "Add panel"
  2. Enter query:
sum(rate(django_http_responses_total_by_status_total{status=~"5.."}[5m]))
  1. Set title: "5xx Error Rate"
  2. Set unit: "errors/sec"
  3. Click "Apply"

Step 5: Save Dashboard

  1. Click "Save dashboard" (disk icon)
  2. Name: "My Django Dashboard"
  3. Click "Save"

🎨 Dashboard Best Practices

Layout Tips

┌─────────────────────────────────────────────────────────────────┐
│  Row 1: Key Metrics (Request Rate, Latency, Errors)            │
├─────────────────────────────────────────────────────────────────┤
│  Row 2: Resource Metrics (CPU, Memory, Disk)                   │
├─────────────────────────────────────────────────────────────────┤
│  Row 3: Database Metrics (Query Time, Connections)             │
├─────────────────────────────────────────────────────────────────┤
│  Row 4: Logs (Recent errors, Recent requests)                  │
└─────────────────────────────────────────────────────────────────┘

Panel Types

Type Best For
Graph Time-series data (rates, latencies)
Stat Single value (current CPU, total requests)
Table Detailed data (top endpoints)
Logs Log entries

🔬 Hands-On Exercise

Create a Complete Dashboard

  1. Request Rate - sum(rate(django_http_requests_total[5m]))
  2. P95 Latency - histogram_quantile(0.95, sum(rate(django_http_requests_latency_seconds_bucket[5m])) by (le))
  3. Error Rate - sum(rate(django_http_responses_total_by_status_total{status=~"5.."}[5m]))
  4. CPU Usage - 100 - (avg(rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)
  5. Memory Usage - (1 - (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes)) * 100
  6. Recent Logs - {app="django", level="ERROR"}

🔧 Troubleshooting

No data in panels

Check datasource:

  1. Go to Configuration → Data Sources
  2. Verify Prometheus URL: http://obs-prometheus:9090
  3. Click "Save & Test"

Check if metrics exist:

curl http://localhost:9090/api/v1/query?query=up
Loki datasource not working

Verify Loki datasource:

  1. Go to Configuration → Data Sources
  2. Click Loki datasource
  3. URL should be: http://obs-loki:3100
  4. Click "Save & Test"

📚 What You Learned

In this step, you learned:

  • ✅ How to access and login to Grafana
  • ✅ How to use Explore for ad-hoc queries
  • ✅ How to create dashboards
  • ✅ How to add panels for metrics and logs
  • ✅ Dashboard best practices

🎉 Tutorial Complete!

Congratulations! You've completed the Django Observability tutorial. You now know how to:

  1. ✅ Set up the observability stack
  2. ✅ Monitor metrics with Prometheus
  3. ✅ Aggregate logs with Loki
  4. ✅ Set up alerts with Alertmanager
  5. ✅ Visualize data with Grafana

🎓 Continue Learning

Explore the detailed module documentation:

Module What You'll Learn
Django App Deep dive into metrics & logging
Prometheus Advanced PromQL queries
Grafana Advanced dashboards
Loki Advanced LogQL queries
Alertmanager Advanced alert routing
MCP Server AI integration

Thank you for following the tutorial! 🎉

View All Modules →