Docker Cleanup & Pruning

Every container you run, every image you pull, every build you run, it all piles up quietly in the background. Left unchecked, it's not unusual to find Docker eating 20 to 30GB of disk space after a few weeks of active development. Therefore, it becomes important to clean it up safely.

See What's Taking Up Space First

Before deleting anything, check what you actually have:

# Show running containers
docker ps

# Show ALL containers, including stopped ones
docker ps -a

# Count how many stopped containers are sitting around
docker ps -aq -f status=exited | wc -l

# Disk usage breakdown across all Docker objects
docker system df

docker system df output:

The RECLAIMABLE column is your target, it shows exactly how much space you can get back without affecting anything currently in use.

Removing Containers

# Remove a stopped container
docker rm my-container

# Remove multiple at once
docker rm web db cache

# Force remove a running container (in background, it stops first, then deletes)
docker rm -f my-container

# Remove a container and its anonymous volumes
docker rm -v my-container

You can't remove a running container without -f. Either stop it first with docker stop, or force-remove it with docker rm -f.

Remove All Stopped Containers at Once

# Remove all stopped containers (It prompts for confirmation)
docker container prune

# Skip the confirmation
docker container prune -f

# Only remove containers stopped more than 24 hours ago
docker container prune --filter "until=24h"

--rm

The most effective way to avoid container buildup is to never let them accumulate in the first place. Add --rm to any container you don't need to keep. Use --rm for one-off commands, CI tasks, data migrations, quick tests or anything you won't need to revisit. It's the single best cleanup habit you can build.

# Container is automatically deleted the moment it exits
docker run --rm alpine echo "done"
docker run --rm -it ubuntu bash

Removing Images

Old and unused images are usually the biggest disk consumers:

# Remove a specific image
docker rmi nginx:1.24

# Remove multiple images
docker rmi nginx:1.24 node:18

# Force remove (even if a stopped container references it)
docker rmi -f old-image:v1

# Remove dangling images (untagged layers left over from builds)
docker image prune

# Remove ALL images not used by any container
docker image prune -a

# Skip confirmation
docker image prune -a -f

Dangling images are the untagged <none> images you see in docker images, they're leftover intermediate layers from previous builds. Safe to delete.

Removing Volumes

# Remove a specific volume
docker volume rm mydata

# Remove all volumes not attached to any container
docker volume prune

# Skip confirmation
docker volume prune -f

docker volume prune permanently deletes data. Always run docker volume ls first to make sure you're not about to wipe something important. There's no undo.

Removing Networks

# Remove a specific network
docker network rm my-network

# Remove all unused networks
docker network prune

Unused networks are safe to prune as they hold no data, just configuration.

Build Cache Cleanup

Build cache is almost always the biggest and safest thing to delete:

# Remove all build cache
docker builder prune

# Remove cache older than 48 hours
docker builder prune --filter "until=48h"

# Skip confirmation
docker builder prune -f

Build cache is 100% safe to delete. The only consequence is that your next build starts from scratch and takes longer, but subsequent builds will cache again normally.

The Nuclear Option: docker system prune

When you want to clean everything at once:

# Remove stopped containers, dangling images, unused networks, build cache
docker system prune

# Also remove unused volumes
docker system prune --volumes

# Also remove ALL unused images (not just dangling ones)
docker system prune -a

# Full cleanup, no prompts
docker system prune -a --volumes -f

Don't run docker system prune -a --volumes unless you're sure. It removes every container, image, and volume not currently in use, including volumes with data you might still need.

A Practical Cleanup Routine

Daily during active development:

docker container prune -f    # Stopped containers gone
docker builder prune -f      # Build cache cleared

Weekly:

docker image prune -a -f     # Unused images removed
docker volume prune -f       # Orphaned volumes cleared

Quick Reference

Key Takeaways

• docker system df is always your first move as it shows exactly how much space each type of Docker object is using and how much is reclaimable
• Use --rm on docker run for any container you won't need again.
• Build cache is always safe to delete with docker builder prune
• docker volume prune is destructive, data in orphaned volumes is gone permanently; always check docker volume ls first
• docker system prune -a --volumes is the option used only when you're certain nothing currently unused is worth keeping
• A regular cleanup routine (daily containers + cache, weekly images + volumes) keeps disk usage manageable before it becomes a problem