How do you clean up Docker resources (stopped containers, unused images)?

Direct Answer

Start with docker system df, then remove only confirmed unused objects. docker container prune removes stopped containers; docker image prune removes dangling images by default (-a broadens it); docker network prune removes unused networks; and build-cache/volume cleanup is separate. docker system prune combines several categories but does not remove named volumes unless explicitly requested.

Code Example

// Check Docker disk usage first
docker system df

// Remove all stopped containers, unused networks, dangling images, and build cache
docker system prune

// More aggressive: include unused images (not just dangling)
docker system prune -a

// Remove only stopped containers
docker container prune

// Remove only unused images
docker image prune

// Remove specific container
docker rm order-api

// Remove specific image
docker rmi order-service:v1

// Remove all stopped containers (manual approach)
docker rm $(docker ps -a -q -f status=exited)

// Force remove running container
docker rm -f order-api

// Remove image even if containers reference it
docker rmi -f order-service:v1

Common Mistakes & Pitfalls

• Not knowing Docker caches everything: Freshers run docker build dozens of times during a feature, never realizing that each failed build leaves layers behind. Weeks later, Docker uses 25GB and they think it's a bug. Running docker system df reveals the truth — build cache alone can hit 15GB.
• Confusing docker rm and docker rmi: Trying to delete an image with docker rm order-service:v1 fails with a confusing error. Remember: rm for containers (running instances), rmi for images (templates). The names are frustratingly similar.
• Pruning without checking first: Running docker system prune -a blindly deletes all unused images, forcing you to rebuild or re-download everything on next use. Always run docker system df first to see what you're about to delete and whether it's worth the rebuild cost.