| Summary: | sys-process/fcron refuses to stop with --pidfile in start-stop-daemon | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Steve Dibb (RETIRED) <beandog> |
| Component: | New packages | Assignee: | Thomas Deutschmann (RETIRED) <whissi> |
| Status: | RESOLVED NEEDINFO | ||
| Severity: | normal | CC: | barzog, cron-bugs+disabled, staff |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
Define "multiple PIDs" please.. cron tasks running? (In reply to comment #1) > Define "multiple PIDs" please.. cron tasks running? Nope, it just creates them sometimes. # ps axf | grep fcron 5804 pts/0 S+ 0:00 \_ grep --colour=auto fcron 5357 ? Ss 0:00 /usr/sbin/fcron -c /etc/fcron/fcron.conf 5710 ? S 0:00 \_ /usr/sbin/fcron -c /etc/fcron/fcron.conf # pidof fcron 5710 5357 # cat /var/run/fcron.pid 5357 Check for defunct processes, please. Those really look like the child fcron spawns to execute tasks, and those are correctly not TERM'd by s-s-d. (In reply to comment #4) > Check for defunct processes, please. > > Those really look like the child fcron spawns to execute tasks, and those are > correctly not TERM'd by s-s-d. Hmm, in that case, there may be some runaway cron jobs. I'll keep an eye on the logs and do some research. I do see this: Oct 24 14:10:00 localhost fcron[5357]: process already running: systab's test -x /usr/sbin/run-crons && /usr/sbin/run-crons Oct 24 14:20:00 localhost fcron[5357]: process already running: systab's test -x /usr/sbin/run-crons && /usr/sbin/run-crons Oct 24 14:30:00 localhost fcron[5357]: process already running: systab's test -x /usr/sbin/run-crons && /usr/sbin/run-crons Oct 24 14:40:00 localhost fcron[5357]: process already running: systab's test -x /usr/sbin/run-crons && /usr/sbin/run-crons *** Bug 444302 has been marked as a duplicate of this bug. *** It seems in my case the problem is with systat: Jan 29 09:00:00 peerbox-r2 fcron[7883]: Job test -x /usr/sbin/run-crons && /usr/sbin/run-crons started for user systab (pid 7884) Jan 29 09:00:00 peerbox-r2 run-crons[7890]: (root) CMD (/etc/cron.hourly/sysstat) Jan 29 09:10:00 peerbox-r2 fcron[7815]: process already running: systab's test -x /usr/sbin/run-crons && /usr/sbin/run-crons root 7815 0.0 0.0 14496 748 ? Ss 08:37 0:00 /usr/libexec/fcron -c /etc/fcron/fcron.conf root 7883 0.0 0.0 52404 2016 ? S 08:59 0:00 /usr/libexec/fcron -c /etc/fcron/fcron.conf root 7884 0.0 0.0 0 0 ? Zs 08:59 0:00 [bash] <defunct> /etc/cron.hourly/sysstat: /usr/lib64/sa/sa1 600 6 & Are you running multiple fcron instances? There were some problemes with the runscript which didn't return the pidfile all the time. Can you please retry with =sys-process/fcron-3.2.1? See https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=531e27c45e1f413da44d8a51ece0f59e454586fb Please re-open if you are still see this with fcron-3.2.1. |
Anytime fcron is running more than one process (multiple PIDs), it refuses to stop using start-stop-daemon --pidfile ${FCRON_PIDFILE} However, using -n fcron works great. Lemme know if you need me to post any info. I'd love to patch the init script. This happens all the time, on every box I maintain.