Skip to content


GitHub /


Starting the container

docker run --rm \
    --init \
    -v /<host_folder_branch_1>:/branch_1 \
    -v /<host_folder_branch_2>:/branch_2 \
    -v /<host_folder_mountpoint>:/mountpoint:shared \
    --cap-add SYS_ADMIN \
    --device /dev/fuse \ -o allow_other /branch_1 /branch_2 /mountpoint

The default ENTRYPOINT is mergerfs -f.


Tag Description Version Commit SHA Commit Message Last Updated Age


The above table reflects the current state of the Git repository. The docker repository is mainly in sync, unless builds fail or commits are pushed for which no builds are started (usually indicated with [skip ci] in the commit message). You can click on the tag's name to go to its corresponding branch on GitHub, clicking the commit sha brings you to the git diff of that commit.

Using the mergerfs mount on the host

By setting the bind-propagation to shared on the volume mountpoint, like this -v /data/mountpoint:/mountpoint:shared, you are able to access the mount from the host. If you want to use this mount in another container, the best solution is to create a volume on the parent folder of that mount with bind-propagation set to slave. For example, -v /data:/data:slave (/data on the host, would contain the previously created volume mountpoint). Doing it like this will ensure that when the container creating the mount restarts, the other containers using that mount will recover and keep working.

Extra docker privileges

On some systems you'll also need the following privileges.

--security-opt apparmor:unconfined