Skip to main content
Lesson 07 of 9

Docker Storage and Volumes

Persist data beyond container lifetimes — named volumes, bind mounts, and database examples.

Containers Are Ephemeral

When you remove a container, everything written to its filesystem disappears with it. Imagine running PostgreSQL in Docker, inserting a week’s worth of records, then running docker rm -f my-db. Without external storage, that data is gone — the database files lived inside the container layer and died when the container did.

Volumes solve this by storing data outside the container’s writable layer, so it survives restarts, rebuilds, and removals.

Named Volumes vs Bind Mounts

Both attach external storage to a container, but they work differently:

Named volume Bind mount
Managed by Docker (created with docker volume create) You (a path on the host filesystem)
Good for Database data, app state, anything Docker should manage Live code reload during development, config files you edit by hand
Location Docker-managed directory (usually under /var/lib/docker/volumes/) Any host path you specify, e.g. ./data:/data
Portability Moves with the volume name across machines via backup/restore Tied to a specific host path — breaks if the path does not exist elsewhere

Re-run a container mapped to the same volume and your previous data comes back exactly as you left it.

Persisting Database Data

Here is a Compose-style snippet that keeps PostgreSQL data in a named volume:

services:
  database:
    image: postgres:16
    environment:
      POSTGRES_PASSWORD: secret
      POSTGRES_DB: myapp
    volumes:
      - pgdata:/var/lib/postgresql/data

volumes:
  pgdata:

Docker creates the pgdata volume on first run. PostgreSQL writes to /var/lib/postgresql/data inside the container, but the files live in the volume. Stop the container, start a fresh one with the same volume — all tables and rows are still there.

Why Volumes Matter

  • Data persistence — your app’s state survives container restarts and redeployments instead of resetting every time.
  • Host isolation and performance — Docker manages volume storage separately from the container layer, avoiding filesystem overhead on the writable layer.
  • Safe data sharing — multiple containers can mount the same volume read-only or read-write without copying files between containers.
  • Direct Docker management — use docker volume ls, docker volume inspect, and docker volume rm to manage storage without hunting through host directories.

Key Takeaway Container filesystems are temporary — volumes are not. Use named volumes for data Docker should manage (databases, uploads), bind mounts when you need a specific host path (local dev). Map the same volume on every run and your data persists.