Lesson · Hinglish · ⏱ 7 min padhne mein

📦 Docker: app ko pack karna taaki kahin bhi chale

Q: Docker mainly kaunsi problem solve karta hai?

'Mere laptop pe chalta hai, server pe nahi'. Containers exact environment pack karte hain, isliye app har jagah ek jaisi chalti hai.

🏠 Asli zindagi ki misaal

Ghar shift karte waqt aap kitchen ka saara saamaan (bartan, masaale, gas regulator) ek label wale box mein pack karte ho, taaki naye ghar mein bhi kaam kare. Docker container aapki app ko uske exact Python version aur libraries ke saath pack karta hai, taaki woh laptop, server ya cloud pe ek jaisi chale. Ab 'mere laptop pe to chal raha tha' wali problem nahi.

💡 Aasaan shabdon mein

Dockerfile ek recipe hai: base image se shuru karo (jaise python:3.12-slim), apna code copy karo, requirements install karo, aur chalane wali command set karo. 'docker build' recipe ko image mein badalta hai, ek frozen package, aur 'docker run' usse container start karta hai. Images ko registry pe push karke kisi bhi server pe pull kar sakte ho. Model APIs deploy karne ki standard unit containers hi hain, aur Kubernetes containers ko bade scale pe chalata hai. Images chhoti rakho, aur secrets kabhi image ke andar mat daalo.

🎬 Dekho kaise kaam karta hai

Aapka code, exact Python version aur libraries ek image mein pack hote hain, isliye container laptop, server ya cloud pe ek jaisa chalta hai.

🪜 Step by step

Dockerfile likho: slim Python image FROM se shuru, requirements.txt COPY, pip install RUN, apna code COPY, CMD set karo.
Image build karo: docker build -t my-model:1.0 .
Container chalao: docker run -p 8000:8000 my-model:1.0, phir localhost:8000 kholo.
Configuration aur secrets environment variables se do, image mein band mat karo.
Image ko registry pe push karo taaki servers aur Kubernetes bilkul wahi cheez pull kar sakein.

🔑 Zaroori shabd

Image: App aur uske environment ka packed, read-only snapshot.
Container: Image ka chalta hua instance.
Registry: Images ka storage, jaise Docker Hub ya cloud registry.

👀 Example dekho

dockerfile

FROM python:3.12-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt
COPY . .
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8000"]

# docker build -t price-api . && docker run -p 8000:8000 price-api

Chhe lines jo aapki model API ko portable bana deti hain.

⚠️ Aam galtiyaan jinse bachna hai

Build tools aur data andar wali bahut badi images. Slim bases, .dockerignore aur multi stage builds use karo.
Production mein 'latest' tag use karna. Versions pin karo taaki pata ho exactly kya chal raha hai.

✅ Jaldi se check karo

Docker mainly kaunsi problem solve karta hai?

Models ko zyaada accurate banaana
'Mere laptop pe chalta hai, server pe nahi'
SQL likhna

Jawaab dekho

B. 'Mere laptop pe chalta hai, server pe nahi': Containers exact environment pack karte hain, isliye app har jagah ek jaisi chalti hai.

🙋 Seekhne wale yeh poochhte hain

Kya Docker virtual machine jaisa hi hai?

Nahi. Containers host ke operating system kernel ko share karte hain, isliye bahut halke hain aur seconds mein start hote hain.

Kya containers GPU use kar sakte hain?

Haan, NVIDIA container toolkit aur CUDA base images ke saath.

📌 Yaad rakho

Dockerfile (recipe) → image (package) → container (chalti app).

📖 Is lesson ke shabd (AI dictionary)

Poora course, aapke roadmap ke hisaab se

Progress tracking, quick checks aur step-by-step order ke saath. Yeh lesson in courses ka hissa hai:

ML Engineer · MLOps / AI Platform Engineer

Course mein seekho