Ed Swierk wrote: > I found that adding RemainAfterExit=yes to the service file prevents > this from happening. But nothing in the systemd documentation (nor in > the code, to the extent I understand it) indicates that this is > necessary for oneshot services.
This has been a known issue for some services, and has been fixed for at least some of them upstream by adding RemainAfterExit. Without RemainAfterExit, when another service is started and depends on the oneshot service (directly or indirectly, typically by depending on something like basic.target which depends on the oneshot), systemd will not consider the oneshot to be already active if it has finished executing and exited, and it will thus be started again. Is this really v208? systemd-random-seed.service in v208 DOES have RemainAfterExit=yes. The version in v204 did trigger this issue, but it's set up differently and the part which could trigger is named systemd-random-seed-load.service. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
