Skip to content


Repository files navigation



Build Status GitHub tag Pulls Stars

Enhanced Docker image for Radicale, the CalDAV/CardDAV server.

Table of contents


  • ๐Ÿ” Secured: the container is read-only, with only access to its data dir, and without extraneous privileges
  • ๐Ÿ”ฅ Safe: run as a normal user (not root)
  • ๐Ÿ—๏ธ Multi-architecture: run on amd64 and arm64
  • โœจ Batteries included: git and ssh included for versioning and Pytz/tz-data for proper timezone handling


๐Ÿ“ƒ See

Latest version

latest tag


Option 1: Basic instruction

docker run -d --name radicale \
    -p 5232:5232 \
    -v ./data:/data \

Option 2: Recommended, Production-grade instruction (secured, safe...) ๐Ÿš€

This is the most secured instruction:

docker run -d --name radicale \
    -p \
    --init \
    --read-only \
    --security-opt="no-new-privileges:true" \
    --cap-drop ALL \
    --cap-add CHOWN \
    --cap-add SETUID \
    --cap-add SETGID \
    --cap-add KILL \
    --pids-limit 50 \
    --memory 256M \
    --health-cmd="curl --fail http://localhost:5232 || exit 1" \
    --health-interval=30s \
    --health-retries=3 \
    -v ./data:/data \

A Docker compose file is included.

Note on capabilities:

  • CHOWN is used to restore the permission of the data directory. Remove this if you do not need the chown to be run (see below)
  • SETUID and SETGID are used to run radicale as the less privileged radicale user (with su-exec), and are required.
  • KILL is to allow Radicale to exit, and is required.

Custom configuration

To customize Radicale configuration, first get the config file:


  1. create a config directory (eg. mkdir -p /my_custom_config_directory)
  2. copy your config file into the config folder (e.g. cp config /my_custom_config_directory/config)
  3. mount your custom config volume when running the container: -v /my_custom_config_directory:/config:ro. The :ro at the end make the volume read-only, and is more secured.

Authentication configuration

This section shows a basic example of configuring authentication for Radicale using htpasswd with bcrypt algorithm.
To learn more, refer to the official Radicale document.

First, we need to configure Radicale to use htpasswd authentication and specify htpasswd file's location.
Create a config file inside the config directory (resulting in the path config/config).

hosts =

type = htpasswd
htpasswd_filename = /config/users
htpasswd_encryption = bcrypt

filesystem_folder = /data/collections

Next, create a users file inside the config directory (resulting in the path config/users).
Each line contains the username and bcrypt-hashed password, separated by a colon (:).


Finally, create and run the container using the appropriate volume mount. In this example, both files are stored in the same directory (./config).

docker run -d --name radicale tomsquest/docker-radicale \
    -p 5232:5232 \
    -v ./data:/data \
    -v ./config:/config \

Volumes versus Bind-Mounts

This section is related to the error message chown: /data: Permission denied.

With Docker volumes, and not bind-mounts like shown in the examples above, you may need to disable the container trying to make the data directory writable.

This is done with the TAKE_FILE_OWNERSHIP environment variable.
The variable will tell the container to perform or skip the chown instruction.
The default value is true: the container will try to make the data directory writable to the radicale user.

To disable the chown, declare the variable like this:

docker run -d --name radicale tomsquest/docker-radicale \
    -e "TAKE_FILE_OWNERSHIP=false"

Running with Docker compose

A Docker compose file is included.
It can also be extended.
Make sure you have Docker compose version 2 or higher.


Docker will automatically select the correct image type for your architecture, whether it is amd64 or arm64.

Extending the image

The image is extendable, as per Docker image architecture. You need to create your own Dockerfile.

For example, here is how to add RadicaleIMAP (authenticate by email) and RadicaleInfCloud (an alternative UI) to the image.

Please note that the radicale-imap plugin is not compatible with Radicale 3.0 anymore!

First, create a Dockerfile.extended (pick the name you want) with this content:

FROM tomsquest/docker-radicale

RUN /venv/bin/pip install git+
RUN /venv/bin/pip install git+

Then, build and run it:

docker build -t radicale-extended -f Dockerfile.extended .
docker run --name radicale-extended -p 5232:5232 radicale-extended

Versioning with Git

Radicale supports a hook which is executed after each change to the CalDAV/CardDAV files. This hook can be used to keep a versions of your CalDAV/CardDAV files through git.

This image provides git to support this feature.

Refer to the official documentation of Radicale for the details.

Custom User/Group ID for the data volume

You will certainly mount a volume to keep Radicale data between restart/upgrade of the container. But sharing files from the host and the container can be problematic. The reason is that radicale user in the container does not match the user running the container on the host.

To solve this, this image offers four options (see below for details):

  • Option 0: Do nothing, permission will be fixed by the container itself
  • Option 1: Create a user/group with id 2999 on the host
  • Option 2: Force the user/group ids on docker run
  • Option 3: Build the image with a custom user/group

Option 0: Do nothing, permission will be fixed by the container itself

When running the container with a /data volume (e.g. -v ./data:/data), the container entrypoint will automatically fix the permissions on /data.

This option is OK, but not optimal:

  • Ok for the container, as inside the container, the radicale user can read and write its data
  • But on the host, the data directory will then be owned by the user/group 2999:2999

Option 1: Create a user/group with id 2999 on the host

The image creates a user and a group with id 2999 in the container.
You can create a user/group on your host matching this id.


# On your host
sudo addgroup --gid 2999 radicale
sudo adduser --gid 2999 --uid 2999 --shell /bin/false --disabled-password --no-create-home radicale

Option 2: Force the user/group ids on docker run

The user and group ids used in the container can be overridden when the container is run.
This is done with the UID and GID env variables, e.g. docker run -e UID=123 -e GID=456 ....
This will force all operations to be run with this UID/GID.

โš ๏ธ The --read-only run flag cannot be used in this case. Using custom UID/GID tries to modify the filesystem at runtime but this is made impossible by the --read-only flag.

Option 3: Build the image with a custom user/group

You can build the image with custom user and group ids and still use the --read-only flag.
But you will have to clone this repo, do a local build and keep up with changes of this image.

Usage: docker build --build-arg=BUILD_UID=5000 --build-arg=BUILD_GID=5001 ....

BUILD_UID and BUILD_GID are also supported as environment variables to work around a problem on some Synology NAS. See this PR#68.


The image is tagged with this scheme:

Version number = Architecture + '.' + Radicale version + '.' + This image increment number


  • tomsquest/docker-radicale:amd64.
  • tomsquest/docker-radicale:arm64.

The last number is ours, and it is incremented on new release. For example, made the /config readonly (this is specific to this image).

Running with Podman

Two users have given the instructions they used to run the image with Podman:

Running behind Caddy

Caddy is sitting in front of all my self-hosted services, like Radicale.
It brings https and security (basic authentication).

The following Caddyfile works for me. Note that I don't use Radicale authentication, I have only one user. {

    basicauth {
        tom pas$w0rd


To run the tests:

  1. pip install pipenv
  2. pipenv install -d
  3. pipenv run pytest -v


  1. Create a Git tag, e.g., push it and the CI will build the images and publish them on Docker hub
  2. Update the latest tag
  3. Create release on GitHub (Draft a new release > pick the tag > Generate release notes > Publish release)
  4. Update

Example instructions :

# Update local tags
git fetch --all --tags
# Create tag
TAG= && git tag $TAG && git push origin $TAG
# Update latest tag
git push --delete origin latest && git tag -d latest && git tag latest && git push origin latest