jburks725/docker-bind repository overview

⁠Dockerized BIND

This is a Dockerized BIND installation intended for serving as an internal DNS cache. It also allows serving of internal zones authoritatively, if desired.

⁠Getting/Building The Container

If you just want a DNS cache, you can simply pull jburks725/bind:latest and run it as below. If you want to serve your own internal zones, you have two options.

⁠Zone File Naming Convention

The start script in this container will automatically add any zone files it finds in the /zones directory as master zones in BIND. To accomplish this, your zone files must be named the same as the canonical name of the zone to serve, including the trailing dot.

• To serve a reverse zone for 192.168.0.0/16: 168.192.in-addr.arpa.
• To serve a forward zone for foo.local: foo.local.

Structure of BIND zone files is beyond the scope of this README. There is a good reference here: http://www.zytrax.com/books/dns/ch6/mydomain.html⁠

Note: The zones.rfc1918 file in this image is configured to allow self- hosting of 192.168.in-addr.arpa. If you want to use a different RFC-1918 reverse zone, you'll need to clone the repo and modify zones.rfc1918 to meet your needs. The original file from the bind9 package is included in the repo as zones.rfc1918.orig.

⁠Internal Zones - Completely Static

This method allows you to set up static internal zones. To modify these zones, you will need to rebuild your image and re-run your container.

1. Clone down the Github repo (https://github.com/jburks725/docker-bind.git⁠)
2. Add one or more BIND zone files to the zones/ directory in the working directory, following the naming convention above
3. Build the image normally: docker build -t yourname/bind .

Note: The zones/ directory contains a .gitignore file that will ignore all contents of the directory, so if you're version controlling this after cloning, you may want to modify the .gitignore.

⁠Internal Zones - Volumes and RNDC Updates

If you'd like to be able to update your internal zones without rebuilding the image every time, you may take advantage of Docker's VOLUME and exec features. The image is already setup with a /zones volume, so you may pull and run the image without modification. You'll just need to mount a directory with zone files (pay attention to the naming convention above) to the /zones location.

⁠Running The Container - DNS Cache or Static Zones

If you've pulled the image and just want to run a cache, or you've built your own image with static zones per the above instructions, you may simply run it, exposing ports 53 and 53/udp, as below.

# If you built your own image, be sure to substitute your image tag below
docker run -d --name=dns -p 53:53 -p 53:53/udp --restart=on-failure \
jburks725/bind:latest

⁠Running The Container - Volumes and RNDC Updates

If you'd like to be able to update your internal zones using the Docker VOLUME method described above, simply mount a directory on your Docker host to the /zones volume on the container:

docker run -d --name=dns -p 53:53 -p 53:53/udp -v /path/to/zonefiles:/zones:ro \
--restart=on-failure jburks725/bind:latest

You may then modify the zones in this directory on your Docker host (being sure to update your serial numbers when you do), then reload the zones using RNDC:

docker exec dns rndc reload

You may check the docker logs dns output to verify that your new zone serial numbers were loaded, or you may dig the SOA record for the zone.

If you want to add new zones to your configuration, you'll need to restart the container rather than doing an rndc reload.