Docker简明教程 (it-ebooks)（Z-Library）

# Docker简明教程 (it-ebooks) — Reading Guide ## 【One-Line Pitch】 A practical, hands-on introduction to Docker for developers and system engineers who want to move from understanding container concepts to confidently building, running, and managing containers with real command-line workflows. If you learn best by doing and need a compact reference rather than a deep theoretical tome, this guide is for you. ## 【Book Arc】 - **Opening (~0%–6%)**: Introduces Docker's origin story (from PaaS to container technology) and explains the core kernel technologies—Linux namespaces (PID, IPC, mnt, UTS, user) and Cgroups—that provide isolation and resource control. Also covers AUFS (Union File System) as the foundation for Docker's layered image model. - **Early (~6%–16%)**: Moves into practical territory: installing Docker on CentOS 7, understanding images vs. containers, and building base images (including a debootstrap example). The Dockerfile is introduced with its structure, best practices, and a detailed parameter reference (FROM, RUN, CMD, ENTRYPOINT, EXPOSE, ENV, ADD, COPY, VOLUME, USER, WORKDIR, ONBUILD). - **Early–Middle (~16%–28%)**: Focuses on container lifecycle management: running containers in foreground/background modes, naming, cleanup, and the essential commands (ps, logs, stop, start, rm, inspect, top). Data management is a major theme—data volumes (-v), data volume containers (--volumes-from), and backup/migration patterns. - **Middle (~28%–44%)**: Covers container networking: port mapping (-p/-P), legacy container linking (--link) with its environment variable and /etc/hosts mechanisms, and the trade-offs between these approaches. Includes practical examples like connecting a web app to a database container. - **Late (~44%–53%)**: Dives into daemon configuration and advanced topics: daemon socket options (unix, tcp, fd), TLS security, storage drivers (aufs, devicemapper, btrfs, zfs, overlay), and registry security (insecure registries, CA certificates). The book closes with case studies, including a multi-stage ZooKeeper cluster setup (standalone → single-host cluster → cross-host cluster). ## 【Key Takeaways】 - **Isolation is the heart of containers** (Early): Linux namespaces (PID, IPC, mnt, UTS, user) and Cgroups provide process, filesystem, hostname, and resource isolation—this is what makes containers lightweight compared to VMs. - **AUFS enables the layered image model** (Early): Docker images are read-only layers stacked via Union FS; containers add a thin writable layer on top. This allows image sharing, fast startup, and efficient storage—the conceptual basis for everything that follows. - **Dockerfile is a declarative build recipe** (Early): Each instruction (FROM, RUN, CMD, etc.) creates a new layer. Understanding the difference between CMD (overridable) and ENTRYPOINT (fixed), and using COPY over ADD for local files, are key to writing clean, reproducible builds. - **Data volumes decouple persistence from container lifecycle** (Early–Middle): Volumes (-v) and volume containers (--volumes-from) let you share and persist data across container restarts and removals. Backup via `docker run --volumes-from ... tar` is a simple, powerful pattern. - **Container linking is a legacy but instructive pattern** (Middle): The --link flag creates private channels between containers via environment variables and /etc/hosts entries. It's largely superseded by user-defined networks, but understanding it clarifies how Docker networking evolved. - **Daemon configuration matters for production** (Late): The daemon (docker -d) supports multiple sockets (unix, tcp, fd), TLS for secure remote access, and pluggable storage drivers—each with trade-offs (e.g., aufs shares memory but isn't in the mainline kernel; devicemapper uses COW snapshots). - **Registry security is non-negotiable** (Late): Use TLS with known CAs for private registries; `--insecure-registry` is a convenience that should be avoided in production. The book's examples show how to configure CA certificates and CIDR-based exceptions. ## 【Reading Tips】 - **Skim the theory, focus on commands**: The first ~6% (namespaces, AUFS) is conceptual—read it once for mental models, but don't get bogged down. The real value is in the command examples from ~9% onward. - **Use the command dictionary as a reference**: The book includes a dictionary-style listing of Docker commands (attach, build, commit, cp, diff, events, etc.). Skim this section and return to it when you need a specific command's syntax. - **Deep-read the data and networking chapters**: Sections on data volumes (4.2) and container linking (4.4) are the most practically useful. Work through the examples step-by-step—create a volume container, link two containers, and inspect the environment variables. - **Watch for version-specific advice**: The book references Docker 1.9.0 and older daemon flags (e.g., `docker -d` instead of `dockerd`). If you're on a modern Docker version, verify commands against current documentation. - **Skip the ZooKeeper case study unless needed**: The final cluster-building example (8.x) is valuable for distributed systems, but it's advanced. Return to it after mastering the basics. ## 【Coverage Limits】 This guide covers the book's core content: Docker fundamentals, image building, container management, data volumes, networking, and daemon configuration. The excerpts do not include detailed coverage of Docker Compose, Kubernetes, or modern container orchestration, nor do they cover the ZooKeeper case study in depth. ##