容器化简介

Containerization packages your app and all its dependencies into one isolated unit — a self-contained box that runs anywhere, consistently.

Before containers there were Virtual Machines. Each VM runs a full OS on top of yours — powerful, but heavy on disk, memory, and startup time.

The classic 'works on my machine' bug comes from environment drift: missing libs, different OS versions, clashing deps. Containers kill that problem.

A container bundles code, runtime, tools, and config so it runs identically everywhere — isolated, lightweight (it shares the host kernel), and portable.

A container is a standard, executable unit that packs code plus every dependency, so the app runs the same way across any environment.

Containers vs VMs: VMs ship a full OS each — heavy and slow to boot. Containers share the host kernel, so they start fast and sip resources.

Why bother with containers? Portability across environments, consistent behavior everywhere, lower resource cost than VMs, and near-instant deploys.

Docker is the go-to platform for building, sharing, and running containers. It's the engine that powers those isolated application boxes.

A Docker image is a read-only, executable package — the blueprint defining exactly what goes in a container: the OS layer, your code, and dependencies.

A container is a running instance of an image. Start an image and you get a container — you can run many isolated ones from a single image.

Time for a quick check on what you've learned about containers!

Recap: containers package apps for consistent, portable runs, stay lighter than VMs by sharing the kernel, and Docker images are blueprints, containers their instances.