this tries to solve two issues:
* the current approach mounts the data directories of apps/volumes individually.
this causes a problem with volume mounts that mount after the container is started i.e not
network time/delay but systemd ordering. With CIFS, the mount is a hostname. This requires
unbound to be running but unbound can only start after docker because it wants to bind to
the docker network. one way to fix is to not start sftp automatically and only start sftp
container in the box code. This results in the sftp container attaching itself of the
directory before mounting and it appears empty. (on the host, the directory will appear
to have mount data!)
* every time apptask runs we keep rebuilding this sftp container. this results in much race.
the fix is: mount the parent directory of apps and volumes. in addition, then any specialized appdata
paths and volume paths are mounted individually. this greatly minimized rebuilding and also since we don't rely
on binding to the mount point itself. the child directories can mount in leisure. this limits the race
issue to only no-op volume mounts.
part of #789
Various notes on mounting:
* The permissions come from the mounted file system and not the mount point.
This means that if we change the perms before mounting, it is overridden by
whatever is in the actual file system.
* uid/gid only works for permission-less file systems
SFTP container notes:
* Assumes that nothing changed if the host path hasn't changed. This means that
if a user changes the disk uuid, reload doesn't work.
* Not sure how/why, but even after unmounting the container can still access the old
mount files (!). With ext4 on disk change or nfs after root path change, the file manager
continues to be able to access the old mounts (despite umount succeeding).
All this led to following changes:
* Remove editing of volumes. Just allow editing username/password.
* edit UI then just also provides a way to re-mount.
* Change mode of mountpoint to be 777 post mounting for ease of use. Otherwise, we have to
make the user do this by ssh. this can always become options later.
we try to add a mount. if it fails, it will revert to the previous mount config.
there was a plan to make this work based on systemd-mount but we hit this bug - https://bugzilla.redhat.com/show_bug.cgi?id=1708996
restart service does not rebuild automatically, we should add a route
for that. we need to figure where to scale services etc if we randomly
create containers like that.
