[RESOLVED] transmission-daemon launches manually but not with systemctl

[oshunluvr]

Using Ubuntu server 20.04 and the Transmission web interface, Initially after the upgrade from 18.04 transmission-daemon would not launch.

I suspected it was a permission issue somewhere because when I read the internet a bit I found several instances where it was mentioned that during the 20.04 upgrade, the file and folder permissions required by transmission-daemon were changed. There wasn't a lot of specific instructions on how to actually solve this. I looked at the folders and files in /etc/transmission-daemon and found several of them owned by "vpn", some by my user, and some by "debian-transmission", the later being a user created for use by transmission-daemon. The files used at run time are set with RW owner only permissions and owned by my user. I also observed the group "sticky bit" active on the folder.

So I did the following:
Purged the transmission specific PPA required by bionic but not by focal. This downgraded me from transmission-daemon 3.0 to 2.94.
Tried the default setting.json and my previous version.
I added debian-transmission to the vpn group and added my user to the debian-transmission group.
I changed ownership/group of the files and folders in /etc/transmission-daemon to debian-transmission.
Reverted to my previous settings,json

I can now manually launch transmission-daemon using this command:

/usr/bin/transmission-daemon -f --log-error -g /etc/transmission-daemon/

However when I attempt to launch it using systemctl, it fails with a useless message:

(code=exited, status=1/FAILURE)

The whole status output:

smith@server:~$ sudo systemctl status transmission-daemon
● transmission-daemon.service - Transmission BitTorrent Daemon
    Loaded: loaded (/lib/systemd/system/transmission-daemon.service; enabled; vendor preset: enabled)
   Drop-In: /etc/systemd/system/transmission-daemon.service.d
            └─local.conf
    Active: inactive (dead) (Result: exit-code) since Tue 2022-03-29 08:42:02 EDT; 2s ago
   Process: 101550 ExecStart=/usr/bin/transmission-daemon -f --log-error -g /etc/transmission-daemon (code=exited, status=1/FAILURE)
  Main PID: 101550 (code=exited, status=1/FAILURE)

Mar 29 08:42:02 server systemd[1]: Stopped Transmission BitTorrent Daemon.

The command I'm using to launch is a duplicate of the command used by systemctl. It will actually show a status of "active" after about 8 seconds then immediately fail again and stay failed. Today I observed the files in the /etc/transmission-daemon had changed back to my user ownership and were rw for owner only again. I changed them back to rw for the group (debian-transmission) but this did not fix the systemctl issue and when I launched it manually again, the settings.json file changed back to no group access.

I've played with the rpc username and file ownership and permissions a bit but I'm missing something, but I don't know what.

[claydoh]

Anything interesting in the logging? journalctl -u transmission-daemon.service
What about running it as the user --guessing that the unit is set to use 'debian-transmission' maybe?

I am by no means an expert in systemd, but I think it may be the service config as opposed to user permissions, per se
systemctl cat transmission-daemon.service
or dig up the file

https://wiki.debian.org/Transmission

Hopefully I have not added confusion.

[oshunluvr]

Ahhh, you're a genius! Or at least give great hints. Remember I said some of the files were owned by the user "vpn" ? So I ran your cat command and saw this:

smith@server:~$ systemctl cat transmission-daemon.service  
# /lib/systemd/system/transmission-daemon.service
[Unit]
Description=Transmission BitTorrent Daemon
After=network.target

[Service]
User=debian-transmission
Type=notify
ExecStart=/usr/bin/transmission-daemon -f --log-error
ExecStop=/bin/kill -s STOP $MAINPID
ExecReload=/bin/kill -s HUP $MAINPID

[Install]
WantedBy=multi-user.target

# /etc/systemd/system/transmission-daemon.service.d/local.conf
[Unit]
After=sys-devices-virtual-net-tun0.device
Wants=sys-devices-virtual-net-tun0.device

[Service]
User=
User=vpn
Group=
Group=vpn

Type=simple

ExecStart=
ExecStart=/usr/bin/transmission-daemon -f --log-error -g /etc/transmission-daemon

Restart=on-failure
RestartSec=10

Which revealed that /etc/systemd/system/transmission-daemon.service.d/local.conf had "vpn" as the owner and group of transmission-daemon. A bit more web searching, and I reconfigured it to use my local user instead of vpn and voilà! I have successfully d/l'd the new KDEneon ISO via transmission.

I guess I assumed that "purge" would start my from scratch or the user configuration change the web was vague about was the "vpn" user. I manage my vpn manually so maybe the transmission devs were automating the VPN use?

Regardless, thanks.