Feat/metrics and monitoring

backend/package.json

@@ -12,7 +12,8 @@
     "express": "^4.18.2",
     "cors": "^2.8.5",
     "body-parser": "^1.20.2",
-    "pg": "^8.11.0"
+    "pg": "^8.11.0",
+    "prom-client": "^15.1.0"
   },
   "devDependencies": {
     "nodemon": "^3.0.1"

backend/src/index.js

@@ -3,11 +3,39 @@ const cors = require('cors')
 const bodyParser = require('body-parser')
 const fs = require('fs')
 const path = require('path')
+const promClient = require('prom-client')
 
 const app = express()
 app.use(cors())
 app.use(bodyParser.json())
 
+const collectDefaultMetrics = promClient.collectDefaultMetrics
+collectDefaultMetrics()
+
+const httpRequestsTotal = new promClient.Counter({
+  name: 'http_requests_total',
+  help: 'Total number of HTTP requests',
+  labelNames: ['method', 'path', 'status']
+})
+
+const httpRequestDuration = new promClient.Histogram({
+  name: 'http_request_duration_seconds',
+  help: 'Duration of HTTP requests in seconds',
+  labelNames: ['method', 'path', 'status'],
+  buckets: [0.01, 0.05, 0.1, 0.5, 1, 2, 5]
+})
+
+app.use((req, res, next) => {
+  const start = Date.now()
+  res.on('finish', () => {
+    const duration = (Date.now() - start) / 1000
+    const path = req.route ? req.route.path : req.path
+    httpRequestsTotal.inc({ method: req.method, path, status: res.statusCode })
+    httpRequestDuration.observe({ method: req.method, path, status: res.statusCode }, duration)
+  })
+  next()
+})
+
 const PORT = process.env.PORT || 3000
 const DATA_PATH = path.join(__dirname, '..', 'data', 'tasks.json')
 
@@ -62,4 +90,13 @@ app.delete('/tasks/:id', (req, res) => {
 
 app.get('/health', (req, res) => res.json({ status: 'ok' }))
 
+app.get('/metrics', async (req, res) => {
+  try {
+    res.set('Content-Type', promClient.register.contentType)
+    res.end(await promClient.register.metrics())
+  } catch (err) {
+    res.status(500).end(err)
+  }
+})
+
 app.listen(PORT, () => console.log(`TaskFlow backend listening on ${PORT}`))

docker-compose.yml

@@ -1,5 +1,3 @@
-version: '3.8'
-
 services:
   db:
     image: postgres:17
@@ -61,9 +59,45 @@ services:
     networks:
       - taskflow-network
 
+  prometheus:
+    image: prom/prometheus:v2.48.0
+    container_name: taskflow-prometheus
+    ports:
+      - "9090:9090"
+    volumes:
+      - ./infrastructure/monitoring/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml:ro
+      - prometheus-data:/prometheus
+    command:
+      - '--config.file=/etc/prometheus/prometheus.yml'
+      - '--storage.tsdb.path=/prometheus'
+      - '--web.enable-lifecycle'
+    networks:
+      - taskflow-network
+
+  grafana:
+    image: grafana/grafana:10.2.0
+    container_name: taskflow-grafana
+    ports:
+      - "3001:3000"
+    environment:
+      - GF_SECURITY_ADMIN_USER=${GRAFANA_ADMIN_USER:-admin}
+      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_ADMIN_PASSWORD:-admin}
+      - GF_USERS_ALLOW_SIGN_UP=false
+    volumes:
+      - ./infrastructure/monitoring/grafana/provisioning:/etc/grafana/provisioning:ro
+      - grafana-data:/var/lib/grafana
+    depends_on:
+      - prometheus
+    networks:
+      - taskflow-network
+
 volumes:
   db-data:
     driver: local
+  prometheus-data:
+    driver: local
+  grafana-data:
+    driver: local
 
 networks:
   taskflow-network: