Closed vassilmladenov closed 2 years ago
@haugene any idea ? @vassilmladenov when you check the id, are you checking this inside the docker ? this looks like some issue with linux nfs mount and docker and not so much a bug with this container..should consider moving this to discussions instead
@pkishino I’m checking the id in the host. The reason I made an issue here is that until a few days ago I was using linuxserver/transmission
on this very mount with PUID/PGID set to 1000 and it was working fine.
I looked at this again and I think the issue is the user that runs the setup script of the container, not the PGID. I did the following:
- /home/vmladenov/transmission-home:/config/transmission-home
TRANSMISSION_HOME=/config/transmission-home
I used /config
because I needed something outside /data
and I saw the container sets up a /config
volume.
On the next run, it got past the setup and the container started properly. Plus, the files get the correct permissions. I've updated the title of the issue. It seems to be the same problem described in #1828
@haugene Can we get the scripts to simply be run by PGID if this is set?
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.
Feel free to re-open this issue if you think it deserves another look.
Hi just wanted to add that I'm seeing this as well and so far not sure how to get around it. Setting the container to privileged mode solved it for me, but I don't want to give the container that much capability if only NET_ADMIN should be required. I have my NFS export set up with root_squash and anonuid=1000, anongid=1000 so I should be able to get things running, but I am stuck in a crash loop
I can confirm this is also an issue. I worked around it by attaching to the docker container, creating a new user, run python3 /etc/transmission/updateSettings.py /etc/transmission/default-settings.json ${TRANSMISSION_HOME}/settings.json
from /etc/transmission/start.sh
, then comment out that line, restart the container and then it worked...obviously not ideal.
Update!
...setting TRANSMISSION_HOME=/etc/transmission/
. The downside is that your transmission settings won't be persistent. It also looks like you'll lose the actual .torrent files too.
I am posting here because I came across this issue as well and would like to share my solution similar to what was proposed above.
Set TRANSMISSION_HOME=/transmission-home
env var. And I map this to a local folder on the docker host -v /home/your-user/transmission:/transmission-home
. This solves any permission issues with the files handled within that folder.
And I keep my /data
directory mapped to a folder on my NFS share, which is where everything is downloaded anyways (/data/completed
& /data/incompleted
).
Some additional info about my set up, in case it might affect anything:
/etc/fstab
with these settings: defaults,_netdev
PUID=1000
& PGID=1000
, which matches my docker user.
Is there a pinned issue for this?
Is there an existing or similar issue/discussion for this?
Is there any comment in the documentation for this?
Is this related to a provider?
Are you using the latest release?
Have you tried using the dev branch latest?
Config used
docker-compose.yml
Current Behavior
/mnt/bittorrent is an NFS mount. The mount is configured without
no_root_squash
on the NFS server.I get this error when I start the program
On the client,
mkdir /mnt/transmission/transmission-home
works butsudo mkdir /mnt/transmission/transmission-home
does not (as expected). I expected the PUID/PGID variables to make the container with the correct permissions.I confirmed
uid=1000(vmladenov) gid=1000(vmladenov)
on the client withid -a
.Expected Behavior
It creates a directory at /mnt/bittorrent/transmission-home and runs correctly
How have you tried to solve the problem?
I can make this work by making the NFS mount
no_root_squash
, but then all of the directories haveroot:root
permissions instead ofvmladenov:vmladenov
.Log output
Environment
Anything else?
No response