opt(deploy): multi-stage frontend (node→nginx), non-root user, healthchecks, resource limits

Frontend:
- Multi-stage build: builder (node:22-alpine) → runner (nginx:alpine)
  replaces vite preview (400MB RAM) with nginx static serving (<20MB RAM)
- New nginx-frontend.conf with gzip + long-term asset cache + SPA fallback

Backend & Discord Gateway:
- Add non-root user (USER app) for container security
- chown app files to avoid permission issues
- Add HEALTHCHECK: backend via /api/health, gateway via kill -0 1

Docker Compose:
- Remove unnecessary backend→gateway depends_on
- Add deploy.resources.limits.memory for all services
- Add healthchecks for all services

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com

infra/docker/Dockerfile.backend

@@ -5,17 +5,20 @@ WORKDIR /app
 # Install pnpm
 RUN npm install -g pnpm
 
+# Create non-root user
+RUN addgroup -S app && adduser -S -G app app
+
 # Copy workspace files
 COPY pnpm-workspace.yaml .
 COPY pnpm-lock.yaml .
 COPY package.json .
 
-# Copy packages (workspace dependencies)
-COPY packages/shared ./packages/shared
-
 # Copy patches (pnpm patchedDependencies)
 COPY patches ./patches
 
+# Copy packages (workspace dependencies)
+COPY packages/shared ./packages/shared
+
 # Copy service
 COPY services/backend ./services/backend
 
@@ -28,8 +31,18 @@ RUN pnpm --filter './packages/shared' run build
 # Build backend
 RUN pnpm --filter './services/backend' run build
 
+# Own dist + node_modules by non-root user
+RUN chown -R app:app /app
+
+# Switch to non-root user
+USER app
+
 # Expose port
-EXPOSE 3001
+EXPOSE 3000
+
+# Healthcheck — backend serves HTTP on port 3000
+HEALTHCHECK --interval=30s --timeout=10s --start-period=15s --retries=3 \
+  CMD wget -qO- http://localhost:3000/api/health || exit 1
 
 # Start backend
 CMD ["node", "services/backend/dist/index.js"]

infra/docker/Dockerfile.discord-gateway

@@ -5,6 +5,9 @@ WORKDIR /app
 # Install pnpm
 RUN npm install -g pnpm
 
+# Create non-root user
+RUN addgroup -S app && adduser -S -G app app
+
 # Copy workspace files
 COPY pnpm-workspace.yaml .
 COPY pnpm-lock.yaml .
@@ -37,5 +40,17 @@ RUN pnpm --filter './services/discord-gateway' run build
 # Create recordings directory
 RUN mkdir -p /app/recordings
 
+# Own everything by non-root user
+RUN chown -R app:app /app
+
+# Switch to non-root user
+USER app
+
+# Expose no HTTP port (gateway is internal-only)
+
+# Healthcheck — verify PID 1 (node) is still running (no HTTP server in this service)
+HEALTHCHECK --interval=30s --timeout=5s --start-period=30s --retries=3 \
+  CMD sh -c "kill -0 1"
+
 # Start discord gateway
 CMD ["node", "services/discord-gateway/dist/index.js"]

infra/docker/Dockerfile.frontend

@@ -1,4 +1,5 @@
-FROM node:22-alpine
+# ---- Builder Stage ----
+FROM node:22-alpine AS builder
 
 ARG VITE_BE_API_URL
 ARG VITE_BE_WS_URL
@@ -25,8 +26,15 @@ RUN pnpm install --frozen-lockfile
 # Build frontend (env vars injected at build time)
 RUN VITE_BE_API_URL=${VITE_BE_API_URL} VITE_BE_WS_URL=${VITE_BE_WS_URL} pnpm --filter './services/frontend' run build
 
-# Expose port
+# ---- Runner Stage ----
+FROM nginx:alpine
+
+# Copy Nginx config
+COPY infra/docker/nginx/nginx-frontend.conf /etc/nginx/conf.d/default.conf
+
+# Copy built static files from builder stage
+COPY --from=builder /app/services/frontend/dist /usr/share/nginx/html
+
 EXPOSE 3000
 
-# Start frontend (production preview)
-CMD ["pnpm", "--filter", "./services/frontend", "run", "preview"]
+CMD ["nginx", "-g", "daemon off;"]

infra/docker/docker-compose.yml

@@ -14,7 +14,15 @@ services:
       - "traefik.http.services.imphenbot.loadbalancer.server.port=80"
     depends_on:
       - backend
-      - frontend
+    healthcheck:
+      test: ["CMD", "nginx", "-t"]
+      interval: 30s
+      timeout: 5s
+      retries: 3
+    deploy:
+      resources:
+        limits:
+          memory: 64M
     networks:
       - app-shared-net
 
@@ -28,8 +36,17 @@ services:
     environment:
       NODE_ENV: production
       WEBSERVER_PORT: 3000
-    depends_on:
-      - discord-gateway
+    # Backend talks to gateway via Redis+Postgres, not directly — no depends_on needed
+    healthcheck:
+      test: ["CMD", "wget", "-qO-", "http://localhost:3000/api/health"]
+      interval: 30s
+      timeout: 10s
+      start_period: 15s
+      retries: 3
+    deploy:
+      resources:
+        limits:
+          memory: 256M
     networks:
       - app-shared-net
 
@@ -44,14 +61,35 @@ services:
       NODE_ENV: production
     volumes:
       - ./recordings:/app/recordings
+    # Gateway has no HTTP server — check if PID 1 (node) is alive
+    healthcheck:
+      test: ["CMD-SHELL", "kill -0 1 || exit 1"]
+      interval: 30s
+      timeout: 5s
+      start_period: 30s
+      retries: 3
+    deploy:
+      resources:
+        limits:
+          memory: 512M
     networks:
       - app-shared-net
 
-  # Frontend Service (React Dashboard)
+  # Frontend Service (React Dashboard) — Nginx serving static files
   frontend:
     image: ghcr.io/${OWNER:-mytheclipse}/bete-frontend:latest
     container_name: imphenbot-frontend
     restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "wget", "-qO-", "http://localhost:3000/"]
+      interval: 30s
+      timeout: 5s
+      start_period: 5s
+      retries: 3
+    deploy:
+      resources:
+        limits:
+          memory: 32M
     networks:
       - app-shared-net
 

infra/docker/nginx/nginx-frontend.conf

@@ -0,0 +1,24 @@
+# Nginx config for serving Vite-built static frontend files
+server {
+    listen 3000;
+    server_name _;
+
+    root /usr/share/nginx/html;
+    index index.html;
+
+    # Gzip compression for faster load times
+    gzip on;
+    gzip_types text/plain text/css application/json application/javascript image/svg+xml;
+    gzip_min_length 256;
+
+    # Cache static assets (JS/CSS hashed filenames)
+    location /assets/ {
+        expires 1y;
+        add_header Cache-Control "public, immutable";
+    }
+
+    # SPA fallback — all non-file routes serve index.html
+    location / {
+        try_files $uri $uri/ /index.html;
+    }
+}