Monitoring with Prometheus and Grafana

1services: 
2  # Collects container metrics
3  cadvisor: 
4    image: gcr.io/cadvisor/cadvisor:latest
5    container_name: cadvisor
6    privileged: true
7    devices: 
8      - /dev/kmsg:/dev/kmsg
9    volumes: 
10      - /:/rootfs:ro
11      - /var/run:/var/run:ro
12      - /sys:/sys:ro
13      - /var/lib/docker/:/var/lib/docker:ro
14    restart: unless-stopped
15
16  # Stores metrics
17  prometheus: 
18    image: prom/prometheus:latest
19    container_name: prometheus
20    volumes: 
21      - ./prometheus.yml:/etc/prometheus/prometheus.yml
22      - prometheus_data:/prometheus
23    command: 
24      - '--config.file=/etc/prometheus/prometheus.yml'
25      - '--storage.tsdb.retention.time=15d'
26    restart: unless-stopped
27
28  # Visualizes metrics
29  grafana: 
30    image: grafana/grafana:latest
31    container_name: grafana
32    environment: 
33      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD}
34      - GF_USERS_ALLOW_SIGN_UP=false
35    volumes: 
36      - grafana_data:/var/lib/grafana
37    ports: 
38      - "3000:3000"
39    restart: unless-stopped
40
41volumes: 
42  prometheus_data: 
43  grafana_data: 

1# prometheus.yml
2global: 
3  scrape_interval: 15s
4  evaluation_interval: 15s
5
6scrape_configs: 
7  # Prometheus itself
8  - job_name: 'prometheus'
9    static_configs: 
10      - targets: ['localhost:9090']
11
12  # Container metrics from cAdvisor
13  - job_name: 'cadvisor'
14    static_configs: 
15      - targets: ['cadvisor:8080']
16
17  # Optional: Host metrics from Node Exporter
18  - job_name: 'node'
19    static_configs: 
20      - targets: ['node-exporter:9100']
21
22  # Optional: Docker daemon metrics
23  - job_name: 'docker'
24    static_configs: 
25      - targets: ['host.docker.internal:9323']

1# 1. Access Grafana at http://your-host:3000
2# 2. Login with admin / your-password
3
4# 3. Add Prometheus data source:
5#    - Go to Configuration > Data Sources
6#    - Add Prometheus
7#    - URL: http://prometheus:9090
8#    - Click "Save & Test"
9
10# 4. Import dashboards:
11#    - Go to Dashboards > Import
12#    - Enter dashboard ID and click Load
13
14# Recommended dashboard IDs:
15# - 193: Docker monitoring (cAdvisor)
16# - 1860: Node Exporter Full (host metrics)
17# - 11074: Docker Dashboard
18# - 14282: cAdvisor + Prometheus
19
20# Or create your own with these queries:
21# Container CPU: rate(container_cpu_usage_seconds_total[1m])
22# Container Memory: container_memory_usage_bytes
23# Container Network: rate(container_network_transmit_bytes_total[1m])

1services: 
2  node-exporter: 
3    image: prom/node-exporter:latest
4    container_name: node-exporter
5    command: 
6      - '--path.rootfs=/host'
7      - '--collector.filesystem.mount-points-exclude=^/(sys|proc|dev|host|etc)($$|/)'
8    volumes: 
9      - /:/host:ro,rslave
10    pid: host
11    restart: unless-stopped
12
13  # Update prometheus.yml to scrape node-exporter:
14  # - job_name: 'node'
15  #   static_configs:
16  #     - targets: ['node-exporter:9100']

1# Grafana alerting (simpler for small setups)
2# 1. Edit a panel in Grafana
3# 2. Go to Alert tab
4# 3. Create alert rule
5# 4. Configure notification channel (email, Slack, etc.)
6
7# Prometheus alerting (more powerful)
8# Add to prometheus.yml:
9rule_files: 
10  - "alerts.yml"
11
12alerting: 
13  alertmanagers: 
14    - static_configs:
15        - targets: ['alertmanager:9093']
16
17---
18# alerts.yml
19groups: 
20  - name: container-alerts
21    rules: 
22      - alert: ContainerDown
23        expr: absent(container_last_seen{name="important-container"})
24        for: 1m
25        labels: 
26          severity: critical
27        annotations: 
28          summary: "Container is down"
29
30      - alert: HighCPU
31        expr: rate(container_cpu_usage_seconds_total[5m]) > 0.9
32        for: 5m
33        labels: 
34          severity: warning
35        annotations: 
36          summary: "Container CPU usage > 90%"
37
38      - alert: HighMemory
39        expr: container_memory_usage_bytes / container_spec_memory_limit_bytes > 0.9
40        for: 5m
41        labels: 
42          severity: warning
43        annotations: 
44          summary: "Container memory usage > 90%"

1services: 
2  uptime-kuma: 
3    image: louislam/uptime-kuma:1
4    container_name: uptime-kuma
5    volumes: 
6      - ./uptime-kuma:/app/data
7    ports: 
8      - "3001:3001"
9    restart: unless-stopped
10
11# Features:
12# - HTTP/HTTPS monitoring
13# - TCP port monitoring
14# - Docker container monitoring
15# - DNS monitoring
16# - Multiple notification channels
17# - Status pages
18# - No configuration files needed
19
20# For Docker monitoring, add monitors for:
21# - Container HTTP endpoints
22# - TCP ports your services expose
23# - Docker socket (advanced)