How do you persist data beyond a container's life with volumes?

Short answer

Mount a volume at the data directory (-v name:/container/path, e.g. -v postgres_data:/var/lib/postgresql); writes there live outside the container and survive stop/rm. Volumes persist until docker volume rm/prune, can be shared across containers, and are inspectable in Desktop's Volumes view.

Detailed explanation

A container's writable layer dies with docker rm — restart a database container and you get an empty database. Volumes are the symlink-like escape hatch: docker volume create log-data then docker run -d -v log-data:/logs … stores everything written to /logs outside the container; Docker auto-creates the volume if missing. The docs' Postgres drill proves it: start docker run --name=db -e POSTGRES_PASSWORD=secret -d -v postgres_data:/var/lib/postgresql postgres:18, docker exec -ti db psql -U postgres to CREATE TABLE + INSERT two rows, docker stop db && docker rm db, then docker run --name=new-db -d -v postgres_data:/var/lib/postgresql postgres:18 and SELECT shows both rows — note POSTGRES_PASSWORD is only needed at bootstrap. Manage growth with docker volume ls, docker volume rm <name> (only when detached), and docker volume prune for all unused; Desktop's Volumes view shows stored data, in-use containers/targets, and export/empty/clone actions. Same volume on multiple containers shares files (log aggregation, pipelines).

Official docs:Docker docs: Persisting container data

Common mistakes

• Expecting data to survive docker rm without a volume.
• Running docker volume rm on an attached volume or forgetting prune lets disk fill silently.