🗓️ 11112024 1430

DOCKER OVERVIEW

What it is:

  • Platform for developing, shipping, running applications in containers
  • Packages applications with all dependencies into standardized units
  • Runs consistently across different environments

Problem it solves:

  • "Works on my machine" syndrome
  • Applications behave differently in dev, test, production due to dependency mismatches, configuration drift, environment differences
  • Docker bundles application with entire runtime environment

Containers vs VMs​

Docker (OS-level virtualization)​

  • Containers share host OS kernel
  • Run in isolated user spaces
  • Lightweight (MBs vs GBs)
  • Fast startup (seconds vs minutes)
  • Efficient resource usage (no duplicate OS overhead)

VMs (hardware virtualization)​

  • Full OS per instance with own kernel
  • Complete isolation
  • Heavier resource overhead

Trade-off:

  • Containers: Less isolation (share kernel), can't run Windows containers on Linux hosts
  • VMs: More isolation but slower and heavier

For kernel-level security isolation or running different OS types, use VMs.

For understanding underlying technology, see virtualization.

Core Components​

Docker Engine​

  • Runtime that builds and runs containers
  • Includes daemon (background service) and CLI

Images and Containers​

Dockerfile​

  • Text file with build instructions
  • Defines base image, copies files, installs dependencies, sets up environment

Docker Registry​

  • Where images live (Docker Hub, private registries, cloud registries)
  • Push to share, pull to download

For multi-container applications, see docker_compose.

When to Use Docker​

Use for:

  • Microservices architecture with isolated services
  • Consistent environments across dev/test/prod
  • Quick deployment and scaling
  • Avoiding dependency conflicts
  • Reproducible CI/CD pipelines

Skip for:

  • Need different OS kernels on same host (use VMs)
  • Require kernel-level security isolation
  • Legacy GUI apps with complex display requirements

Key Trade-offs​

Portability vs Kernel dependency​

  • Containers portable across hosts with same kernel
  • Can't run Linux container on Windows kernel (though Docker Desktop uses VM to bridge this)

Speed vs Isolation​

  • Containers start in seconds, use minimal resources
  • But share kernel = less isolation than VMs
  • Proper configuration critical for security

Stateless design vs Persistence​

  • Containers are ephemeral by design
  • For persistent data, need docker_volumes to separate data from container lifecycle

Common Pitfalls​

DANGER

Storing secrets in images: Secrets baked into images persist in layer history. Anyone with image can extract them. Use environment variables, mounted files, or secrets management instead.

WARNING

Running as root: Containers often default to running as root. Security risk. Always create non-root users in production images.

References​