spegel-org/spegel

Spegel (the Swedish word for mirror) is a tool that acts as a local image mirror inside a Kubernetes cluster. In container-based infrastructure, when a server needs to run a piece of software, it first downloads a container image from an external registry such as Docker Hub. Spegel intercepts those downloads and caches the images within the cluster itself, so the next time the same image is needed it can be fetched from a nearby machine rather than the public internet. The project describes itself as stateless, meaning it does not maintain a persistent database or central storage. It works at the cluster level, so images cached on one node are available to others without manual configuration. The README notes that no explicit setup is needed to start caching: once deployed, Spegel begins caching images from external registries automatically. The practical benefits listed in the README are: faster container startup because images are pulled locally rather than over the internet, resilience against external registry outages, since the local cache can serve images even if Docker Hub or another registry is unavailable, avoidance of rate limits imposed by public registries, and reduced outbound network traffic, which matters in environments where external bandwidth is metered or slow, such as edge deployments. The project is described as evolving, with no guaranteed support or stable API. The developers say help is provided on a best-effort basis, and the primary audience is home lab users and individual contributors rather than large enterprise deployments. Spegel was originally developed at Xenit AB and is now maintained under its own organization. It is licensed under the MIT License and accepts contributions via the guidelines in the repository.