Docker remains a cornerstone of modern dev and DevOps workflows. With Docker Desktop 4.42, released in June 2025, Docker now offers native IPv6, built-in MCP Toolkit, and expanded Model Runner AI capabilities—features that significantly enhance real-world development environments
In this guide, we will build a containerized app featuring persistent data, runtime tuning, monitoring, security, and more—with commands and concepts oriented toward professional use.
1. The Practical Value of These Docker Concepts
- Network parity: IPv6 support eliminates environment discrepancies and future-proofs setups .
- AI workflows: With docker model… commands, you can pull and test LLMs locally from the CLI .
- Service catalog: The integrated MCP Toolkit gives instant access to 100+ tools—like GitHub, MongoDB—isolated in secure containers
- Pro-grade features: This demo includes dynamic resource updates, monitoring, lifecycle control, and cleanups.
2. The Mini-App: Ping Counter
We’ll run a Node.js script that reads and increments a counter in counter.json:
project/
├─ Dockerfile
├─ counter.json ← initializes as {"value":0}
└─ index.js
const fs = require('fs');
const data = JSON.parse(fs.readFileSync('./counter.json'));
data.value += 1;
fs.writeFileSync('./counter.json', JSON.stringify(data));
console.log(`Count is now ${data.value}`);
This simple script demonstrates containers, images, volumes, monitoring, and dynamic tuning.
3. Build the image
Dockerfile:
FROM node:20-alpine
WORKDIR /app
COPY package.json .
RUN npm install --production
COPY . .
CMD ["node", "index.js"]
Build and inspect:
docker build -t ping-counter:1.0 .
docker images
This highlights Docker’s layered approach and the modern docker image command suite.
4. Initial Run (Stateless)
echo '{"value":0}' > counter.json
docker run --name no-store ping-counter:1.0
# → Count is now 1
docker run --name no-store2 ping-counter:1.0
# → Count resets to 1
Shows that containers default to a fresh state every run.
5. Add Persistence via Volumes
docker run -d \
--name counter \
-v counter_data:/app/counter.json \
ping-counter:1.0
docker logs counter # → Count is now 1
docker restart counter && docker logs counter # → Count increments
Volumes allow state persistence beyond container lifecycles—key for real-world apps.
6. Dynamic Runtime Tuning
Modify CPU and memory limits on the fly:
docker update --cpus="0.5" --memory="128m" counter
- This live update doesn't require a container restart
- Flags like
--cpus,--memory, and--cpu-sharesmanage cgroup constraints
7. Live Monitoring
docker stats counter
Monitors runtime resource consumption. For deeper analytics, integrate Docker Desktop’s Resource Usage or Prometheus/OpenTelemetry .
8. Rebuild & Redeploy
If you change index.js to increment by five instead, do:
docker build -t ping-counter:1.1 .
docker stop counter && docker rm counter
docker run -d \
--name counter \
-v counter_data:/app/counter.json \
ping-counter:1.1
docker logs counter # → reflects new count
9. Clean-Up & Security
docker ps -a
docker images
docker volume ls
Clean unused:
docker rm no-store no-store2
docker rmi ping-counter:1.0
docker volume prune
Security best practices:
- Use curated/hardened base images.
- Scan using Docker Scout or Trivy.
- Implement SBOM and image signing
10. Advanced Concepts Snapshot
- Health checks (HEALTHCHECK) to detect container readiness.
- Non-root execution enhances runtime security.
- IPv6 toggling within Docker Desktop network settings
- AI model experimentation: use docker model pull and docker model run.
- MCP Toolkit: run MCP tools via docker mcp or UI
Conclusion
In this guide, we built a real-world Dockerized app from scratch—covering images, containers, persistent volumes, live resource tuning, monitoring, and security best practices. We also touched on modern features like IPv6 support and AI model tooling.
With this foundation, we’re now equipped to apply Docker confidently in professional workflows—and ready to take the next step into orchestration, CI/CD, or advanced container security.
cool!