The bindfs filesystem (based on FUSE) is an overlay filesystem with the capability to remap user and group IDs. The mappings must be given on the bindfs
command line when mounting the filesystem, with the --map
command line option or map
mount option. A mapping of the form 123/456
causes files owned by user 123 to be shown as owned by user 456. A mapping @123/456
does the same for groups. You can use user and group names that are visible to the bindfs process. Beware that only traditional permissions are affected, ACL entries are not remapped.
Example:
bindfs --map=123/456:@staff/nogroup /srv/shared/daat /srv/containers/foo/data
I'm not sure if this would help you since it doesn't do remapping between user namespaces, the remapping is done in the namespace of the bindfs
process.
userns
, but let's make it explicit :)