This post is a follow up to my previous post here
http://lists.alioth.debian.org/pipermail/nut-upsuser/2010-August/006175.html
The UPS NOCOMM condition was triggered again this morning, 10 days after
the last restart. This is the third time it has happened, and every
time it has been after 10 days of uptime - which is starting to look
suspicious. Anyway, I can report that the setuid restart hack that I
posted before works fine. /var/log/syslog shows the old processes
exiting and the new processes starting up successfully. ps -ef | grep
ups returns
root 12694 1 0 Aug30 ? 00:00:00 /sbin/upsmon
nut 12696 12694 0 Aug30 ? 00:00:23 [upsmon] <defunct>
nut 20309 1 0 10:00 ? 00:00:00 /lib/nut/usbhid-ups -a
cp1500avr1
nut 20311 1 0 10:00 ? 00:00:00 /sbin/upsd
root 20313 1 0 10:00 ? 00:00:00 /sbin/upsmon
nut 20314 20313 0 10:00 ? 00:00:00 /sbin/upsmon
showing that the new upsmon has successfully forked into 2 processes
with owners root and nut. I think the root ownership proves that the
setuid worked. As before the old upsmon root process failed to die,
with the old upsmon nut process becoming defunct.
_______________________________________________
Nut-upsuser mailing list
[email protected]
http://lists.alioth.debian.org/mailman/listinfo/nut-upsuser
