feat: Docker deployment (M5)

Multi-stage Dockerfile (deps -> build -> prod-deps -> runtime) that
ships a full production node_modules rather than Next's standalone
output, since standalone tracing is incompatible with a custom server
(noted back in M1). Runs as a non-root user with a read-only rootfs,
dropped capabilities, and tini as PID 1. tsx moves from dev to a real
runtime dependency since the production start script runs server.ts
directly rather than a precompiled bundle. next.config.ts marks
dockerode/systeminformation as serverExternalPackages so Next's
bundler leaves their OS-conditional requires alone.

docker-compose.yml mirrors the sibling stacks' own conventions
(TRAEFIK_HOST/NETWORK_NAME in .env, falcon_network as an external
network, the same traefik.* label shape) so it fits their existing
tooling, plus a new /api/health route and healthcheck.mjs for the
container HEALTHCHECK.

Verified end-to-end with a real `docker build` + `docker compose up`:
non-root/read-only/cap-dropped container boots cleanly, is reachable
by container name from another container on the shared network (as
Traefik would reach it), and the container's own HEALTHCHECK reports
healthy. That run surfaced a real gap - the non-root user got EACCES
on /var/run/docker.sock, since it's owned by root:docker on the host -
fixed via group_add on a DOCKER_GID env var (documented in .env.example
with the command to find it), then re-verified that both docker.sock
access and label-based auto-discovery work correctly under the fix.

.dockerignore

@@ -0,0 +1,7 @@
+node_modules
+.next
+.git
+config
+*.log
+.DS_Store
+.env*

.env.example

@@ -0,0 +1,8 @@
+# Compose-level variables, interpolated into docker-compose.yml's Traefik
+# labels - separate from config/.env, which is for the app's own config.yml.
+TRAEFIK_HOST=pulsenode.example.com
+NETWORK_NAME=falcon_network
+
+# Host's docker group GID, so the non-root container user can read
+# /var/run/docker.sock. Find it with: getent group docker | cut -d: -f3
+DOCKER_GID=999

.gitignore

@@ -32,6 +32,7 @@ yarn-error.log*
 
 # env files (can opt-in for committing if needed)
 .env*
+!/.env.example
 !config/.env.example
 
 # vercel

Dockerfile

@@ -0,0 +1,46 @@
+# output: 'standalone' can't be combined with a custom server (Next traces
+# its own minimal server.js and ignores custom server files), so this image
+# ships a full production node_modules instead of a standalone bundle.
+
+FROM node:22-alpine AS base
+RUN corepack enable
+
+FROM base AS deps
+WORKDIR /app
+COPY package.json pnpm-lock.yaml pnpm-workspace.yaml ./
+RUN pnpm install --frozen-lockfile
+
+FROM base AS build
+WORKDIR /app
+COPY --from=deps /app/node_modules ./node_modules
+COPY . .
+RUN pnpm build
+
+FROM base AS prod-deps
+WORKDIR /app
+COPY package.json pnpm-lock.yaml pnpm-workspace.yaml ./
+RUN pnpm install --prod --frozen-lockfile
+
+FROM node:22-alpine AS runtime
+RUN apk add --no-cache tini
+WORKDIR /app
+ENV NODE_ENV=production
+
+RUN addgroup -S pulsenode && adduser -S pulsenode -G pulsenode
+
+COPY --from=prod-deps /app/node_modules ./node_modules
+COPY --from=build /app/.next ./.next
+COPY --from=build /app/public ./public
+COPY package.json server.ts tsconfig.json next.config.ts healthcheck.mjs ./
+COPY --from=build /app/lib ./lib
+
+RUN mkdir -p /app/config && chown -R pulsenode:pulsenode /app
+
+USER pulsenode
+EXPOSE 3000
+
+HEALTHCHECK --interval=30s --timeout=5s --start-period=15s --retries=3 \
+    CMD ["node", "healthcheck.mjs"]
+
+ENTRYPOINT ["tini", "--"]
+CMD ["node_modules/.bin/tsx", "server.ts"]

app/api/health/route.ts

@@ -0,0 +1,12 @@
+import { configStore } from "@/lib/config/loader";
+
+export const dynamic = "force-dynamic";
+
+export async function GET(): Promise<Response> {
+  try {
+    configStore.get();
+    return Response.json({ status: "ok" });
+  } catch (err) {
+    return Response.json({ status: "error", message: (err as Error).message }, { status: 503 });
+  }
+}

docker-compose.yml

@@ -0,0 +1,52 @@
+services:
+  pulsenode:
+    build: .
+    container_name: pulsenode
+    restart: unless-stopped
+    read_only: true
+    tmpfs:
+      - /tmp
+      - /app/.next/cache
+    cap_drop:
+      - ALL
+    security_opt:
+      - no-new-privileges:true
+    # The container runs as a non-root user, but /var/run/docker.sock is
+    # owned by root:docker on the host - without joining that GID the
+    # container gets EACCES connecting to it. Set DOCKER_GID in .env to
+    # your host's docker group id (`getent group docker | cut -d: -f3`).
+    group_add:
+      - "${DOCKER_GID}"
+    volumes:
+      # Read-only: PulseNode only inspects/lists containers and reads stats,
+      # never execs into or writes to them.
+      - /var/run/docker.sock:/var/run/docker.sock:ro
+      - ./config:/app/config:ro
+    env_file: .env
+    networks:
+      - falcon_network
+    labels:
+      - traefik.enable=true
+      - traefik.docker.network=${NETWORK_NAME}
+      - traefik.http.middlewares.pulsenode-redirect-web-secure.redirectscheme.scheme=https
+      - traefik.http.routers.pulsenode-web.rule=Host(`${TRAEFIK_HOST}`)
+      - traefik.http.routers.pulsenode-web.entrypoints=web
+      - traefik.http.routers.pulsenode-web.middlewares=pulsenode-redirect-web-secure
+      - traefik.http.routers.pulsenode-web-secure.rule=Host(`${TRAEFIK_HOST}`)
+      - traefik.http.routers.pulsenode-web-secure.entrypoints=websecure
+      - traefik.http.routers.pulsenode-web-secure.tls.certresolver=resolver
+      # Verify these entrypoint/middleware/certresolver names against
+      # traefik/dynamic/*.yaml in the stacks repo before deploying - written
+      # to match that repo's own convention, not independently confirmed.
+      - traefik.http.routers.pulsenode-web-secure.middlewares=security-headers@file,no-index@file
+      - traefik.http.services.pulsenode-web-secure.loadbalancer.server.port=3000
+    healthcheck:
+      test: ["CMD", "node", "healthcheck.mjs"]
+      interval: 30s
+      timeout: 5s
+      start_period: 15s
+      retries: 3
+
+networks:
+  falcon_network:
+    external: true

healthcheck.mjs

@@ -0,0 +1,20 @@
+import http from "node:http";
+
+const req = http.request(
+  {
+    host: "127.0.0.1",
+    port: process.env.PORT || 3000,
+    path: "/api/health",
+    timeout: 3000,
+  },
+  (res) => {
+    process.exit(res.statusCode === 200 ? 0 : 1);
+  }
+);
+
+req.on("error", () => process.exit(1));
+req.on("timeout", () => {
+  req.destroy();
+  process.exit(1);
+});
+req.end();

next.config.ts

@@ -1,7 +1,10 @@
 import type { NextConfig } from "next";
 
 const nextConfig: NextConfig = {
-  /* config options here */
+  // dockerode and systeminformation do dynamic/OS-conditional requires
+  // (optional native bindings, platform-specific code paths) that don't
+  // survive Next's bundler well - keep them as plain runtime requires.
+  serverExternalPackages: ["dockerode", "systeminformation"],
 };
 
 export default nextConfig;

package.json

@@ -17,6 +17,7 @@
     "react": "19.2.8",
     "react-dom": "19.2.8",
     "systeminformation": "^5.33.1",
+    "tsx": "^4.23.12",
     "ws": "^8.21.3",
     "yaml": "^2.9.0",
     "zod": "^4.4.3"
@@ -31,7 +32,6 @@
     "eslint": "^9",
     "eslint-config-next": "16.3.1",
     "tailwindcss": "^4",
-    "tsx": "^4.23.12",
     "typescript": "^5"
   },
   "packageManager": "pnpm@11.21.0"

pnpm-lock.yaml

@@ -32,6 +32,9 @@ importers:
       systeminformation:
         specifier: ^5.33.1
         version: 5.33.1
+      tsx:
+        specifier: ^4.23.12
+        version: 4.23.12
       ws:
         specifier: ^8.21.3
         version: 8.21.3
@@ -69,9 +72,6 @@ importers:
       tailwindcss:
         specifier: ^4
         version: 4.3.3
-      tsx:
-        specifier: ^4.23.12
-        version: 4.23.12
       typescript:
         specifier: ^5
         version: 5.9.3