Watchdog
up floh
New Member Posts: 3 ✭
Hi,
Did one of you already enable the hw watchdog?
I have a /dev/watchdog in /dev which looks OK.
I installed the watchdog package, made it point at /dev/watchdog, started the watchdog (/etc/init.d/watchdog, seems it has no unit file) and then kill -9'd the process.
It seems it wasn't restarted via systemd and the system stays up.
I'm not sure if I just missed a BIOS setting or what's wrong with it.
Any ideas?
I'll look back at the BIOS soon, maybe it's just something I missed there.
Did one of you already enable the hw watchdog?
I have a /dev/watchdog in /dev which looks OK.
I installed the watchdog package, made it point at /dev/watchdog, started the watchdog (/etc/init.d/watchdog, seems it has no unit file) and then kill -9'd the process.
It seems it wasn't restarted via systemd and the system stays up.
I'm not sure if I just missed a BIOS setting or what's wrong with it.
root@dhcp106:~# ps -ef | grep watch root 10 2 0 04:41 ? 00:00:00 [watchdog/0] root 11 2 0 04:41 ? 00:00:00 [watchdog/1] root 16 2 0 04:41 ? 00:00:00 [watchdog/2] root 21 2 0 04:41 ? 00:00:00 [watchdog/3] root 2322 1361 0 05:05 pts/0 00:00:00 grep watch root@dhcp106:~# root@dhcp106:~# ls -l /dev/watchdog crw------- 1 root root 10, 130 Aug 17 04:41 /dev/watchdog root@dhcp106:~# grep /dev/watchdog /etc/watchdog.conf watchdog-device = /dev/watchdog root@dhcp106:~# /etc/init.d/watchdog start [ ok ] Starting watchdog (via systemctl): watchdog.service. root@dhcp106:~# ps -ef | grep watch root 10 2 0 04:41 ? 00:00:00 [watchdog/0] root 11 2 0 04:41 ? 00:00:00 [watchdog/1] root 16 2 0 04:41 ? 00:00:00 [watchdog/2] root 21 2 0 04:41 ? 00:00:00 [watchdog/3] root 2403 1 0 05:06 ? 00:00:00 /usr/sbin/watchdog root 2423 1361 0 05:06 pts/0 00:00:00 grep watch root@dhcp106:~# root@dhcp106:~# kill -9 2403 ; sleep 100 ; echo "we r fail" we r fail
Any ideas?
I'll look back at the BIOS soon, maybe it's just something I missed there.
Comments
-
Hi darkfader,
As root, if you run 'touch /dev/watchdog0' it is enough to activate the watchdog timer.
The timer expires by default after a certain amount of time (I thought it should be 60 seconds, but it seems more like 40-50 seconds in the tests I ran).
If you don't "feed it" again before the timer expires, the UP board will automatically reboot.
To feed it, you can run 'echo 1 > /dev/watchdog0'