[ClusterLabs] Q: What is the meaning of "sbd: [19541]: info: Watchdog enabled."

[Jorge Fábregas]

On 01/13/2016 04:34 AM, Ulrich Windl wrote:
> Since an update of sbd in SLES11 SP4 (sbd-1.2.1-0.12.1), I see
> frequent syslog messages like these (grep "Watchdog enabled."
> /var/log/messages):

Hi,

This happened to me as well.  It turned out I was running a monitor on
my SBD resource.  I removed it and everything went back to normal.

I started working with Linux HA with SLES 11 SP4 (so I don't know how
was the behavior before that) but, since you mentioned it started
happening with SP4, then it must be a bug then.

I'm one of those that don't want to see ANYTHING on the logs unless is
completely necessary (no news is good news).

Going back to the original issue,  I guess the proper question is:
Should we run a monitor operation with the SBD resource? I just did a
test on one of my test VMs:  I killed the parent SBD process (kill -9)
and the VM restarted (hard-reset) right away.  I'm not sure who
initiated it (stonithd or pacemaker); it wasn't the watchdog because it
was too fast (I have it set to 15 seconds).  There you go...



Regards,
Jorge