I've been an arch user for more than a year now and this is the first time I've come across a bug that hasn't been solved on the internet and that I can't solve either.
The machine I've had problems on is a VPS running custom glibc but I'm not sure it is relevant to my problem.
Recently I updated systemd to version 197-4 (standard update using pacman -Syu) and then rebooted my machine.
Then systemd-tmpfiles-clean.timer has started running forever causing systemd-journal to use 100% CPU and to fill my hard drive by logging the following lines over and over again:
Jan 17 01:33:51 www systemd: Starting Cleanup of Temporary Directories... Jan 17 01:33:51 www systemd: Failed to start Cleanup of Temporary Directories. Jan 17 01:33:51 www systemd: Starting Cleanup of Temporary Directories... Jan 17 01:33:51 www systemd: systemd-tmpfiles-clean.service start request repeated too quickly, refusing to start. Jan 17 01:33:51 www systemd: Failed to start Cleanup of Temporary Directories.
My VPS admin had to shut my VPS down since it was taking ressources from other virtual machines.
When I stop the timer (as root)
systemctl stop systemd-tmpfiles-clean.timer
everything returns back to normal.
However when I try to run the tmp files clean without the timer:
systemctl start systemd-tmpfiles-clean.service
It is successful
My guess is that this is a timer issue rather than a tmpfiles-clean issue, another reason to believe this is the systemd changelog for this version that indicates a change to calendar timers.
The temporary workaround I found was to deactivate the timer by commenting out the Timer lines in /usr/lib/systemd/system/systemd-tmpfiles-clean.timer
[Timer] #OnBootSec=15min #OnUnitActiveSec=1d
I know this is an ugly hack (that also prevents automatic tmp file cleanup) and was wondering if anyone knew anything about this problem.
Hi, I never found a permanent solution, I've been using the workaround described in my original post since then.