mbentley/apt-cacher-ng repository overview

⁠mbentley/apt-cacher-ng

docker image for apt-cacher-ng based off of debian:trixie / debian:bookworm / debian:bullseye

To pull this image: docker pull mbentley/apt-cacher-ng

⁠Image Tags

⁠Multi-arch Tags

The following tags have multi-arch support for amd64 and arm64 and will automatically pull the correct tag based on your system's architecture:

• latest
• trixie
• bookworm
• bullseye

There are also architecture specific tags if you wish to use an explicit architecture tag:

• latest-amd64, trixie-amd64
• latest-arm64, trixie-amd64
• bookworm-amd64
• bookworm-arm64
• bullseye-amd64
• bullseye-arm64

⁠Date Specific Tags

The latest, trixie, bookworm, and bullseye tags also have unique manifests that are generated daily. These are in the format <tag>-YYYYMMDD (e.g. - latest-20220215) and can be viewed on Docker Hub⁠. Each one of these tags will be generated daily and is essentially a point in time snapshot of the latest tag's manifest that you can pin to if you wish. Please note that these tags will remain available on Docker Hub for 6 months and will not receive security fixes. You will need to update to newer tags as they are published in order to get updated images. If you do not care about specific image digests to pin to, I would suggest just using the latest, trixie, bookworm, or bullseye tags.

⁠Example usage

docker run -d \
  --name apt-cacher-ng \
  -p 3142:3142 \
  -e TZ="US/Eastern" \
  -e PUID=0 \
  -e PGID=0 \
  -v /data/apt-cacher-ng:/var/cache/apt-cacher-ng \
  mbentley/apt-cacher-ng

To change the UID/GID that apt-cacher-ng itself runs as, set PUID and PGID to whatever numerical values you wish.

This image runs apt-cacher-ng, cron, and rsyslogd to ensure that apt-cacher-ng functions properly with scheduled jobs and appropriate logging.

In order to configure a host to make use of apt-cacher-ng on a box, you should create a file on the host /etc/apt/apt.conf with the following lines:

Acquire::http::Proxy "http://<docker-host>:3142";

You can also bypass the apt caching server on a per client basis by using the following syntax in your /etc/apt/apt.conf file:

Acquire::HTTP::Proxy::<repo-url> "DIRECT";

For example:

Acquire::HTTP::Proxy::get.docker.com "DIRECT";
Acquire::HTTP::Proxy::download.virtualbox.org "DIRECT";

Note: The above assumes that you are mapping port 3142 on the docker host and 3142 is accessible from all machines.

You can also update the /etc/apt-cacher-ng/acng.conf and add one or more PassThroughPattern lines to force clients to bypass a repository:

PassThroughPattern: get\.docker\.com
PassThroughPattern: download\.virtualbox\.org

By default, I've enabled a passthrough pattern for anything using port 443.

⁠Drop-in configuration

Instead of bind mounting individual config files, you can mount a directory of *.conf files (plus any files they reference, like backends_* lists) at /etc/apt-cacher-ng/config.d:

docker run -d \
  --name apt-cacher-ng \
  -p 3142:3142 \
  -v /data/apt-cacher-ng:/var/cache/apt-cacher-ng \
  -v /data/apt-cacher-ng-config.d:/etc/apt-cacher-ng/config.d:ro \
  mbentley/apt-cacher-ng

On startup, the files in the top level of config.d are symlinked into /etc/apt-cacher-ng:

• *.conf files are linked as zzzz-dropin-<name>.conf so they are loaded after the default config files and override them
• all other files (e.g. backends_*) are linked with their original name so relative file: references in your *.conf files resolve, since apt-cacher-ng looks them up in /etc/apt-cacher-ng

For example, with config.d/backends_docker containing https://download.docker.com/linux/debian, a config.d/docker.conf can reference it like any other backends file:

Remap-docker: /docker ; file:backends_docker

A drop-in is skipped with a warning if a file with the same name already exists in /etc/apt-cacher-ng (for example, backends_debian and backends_ubuntu ship with the image), so drop-ins can't replace the default files. To use a conflicting file anyway, give it a different name or reference it by absolute path (file:/etc/apt-cacher-ng/config.d/backends_debian). Add or remove files and restart the container to apply the changes.

⁠Drop-in permissions

Unlike the rest of /etc/apt-cacher-ng, the entrypoint does not change the ownership of anything in config.d because it is typically a read-only bind mount and changing ownership would modify the files on the host. Since apt-cacher-ng runs as PUID/PGID (0/0 by default), the drop-in files must be readable by that user or group. If you set a non-root PUID/PGID, either make the files world-readable or have them owned by (or group-readable by) that UID/GID:

chmod 644 /data/apt-cacher-ng-config.d/*

If a drop-in isn't readable by that user, apt-cacher-ng won't be able to load its settings.