What is docker-cleanup?

This project, called Docker Cleanup, automatically frees up disk space on servers by removing old containers and images that are no longer being used. When software containers stop running, they typically linger on disk and accumulate over time. This tool runs on a schedule and clears out that clutter so your servers don't run out of storage. The way it works is straightforward: it runs as a container itself and periodically checks for stopped containers and unused images, removing them after a configurable delay (30 minutes by default). You can adjust how often it cleans up, how long it waits before deleting things, and whether certain images should be protected from deletion. It can also clean up orphaned data volumes if given access to the right directory. By default it runs in a continuous loop, but you can set it to run once and stop. The people who benefit most from this are teams running lots of containers on a single machine, for example, a build server that continuously creates and destroys containers for testing, or a cluster node where many services start and stop throughout the day. Without something like this, leftover containers and outdated image versions gradually fill the disk, which can eventually cause failures or require manual intervention. A Jenkins build server is a concrete example: it rebuilds images frequently, leaving behind untagged versions that pile up over time. Deployment options include running it directly from the command line, or setting it up as a system service that starts automatically and restarts on failure. The README provides configuration examples for systemd-based setups (including CoreOS clusters) and Puppet. Not much else is documented beyond these deployment patterns and the environment variable settings, so users looking for deeper customization or troubleshooting guidance may need to explore the source code.