On 12/7/05, Ed Ravin <[EMAIL PROTECTED]> wrote: > On Wed, Dec 07, 2005 at 01:30:15PM +0100, J Paston wrote: > > Afternoon, > > I have a hosgroup with 20 hosts. One fails. How can the alert script > > know which one failed? Thanks. > > The monitor script is supposed to send the name of the failing host > as the first line of its output (the summary). Please read over the > mon man page - it covers how information is shared between monitors, > mon, and alerts. > > Depending upon what you're doing, you might want to upgrade to the current > mon 1.1 in CVS, since it has a couple of new features for caching of > output from monitors. >
Thanks for the tip. I can't see where it says this in the man page, but now I know. I've noticed that the -u switch is not being sent to the alert script when a service becomes available again. I've confirmed this with the test.alert script. Using mon 1.27 from Debian Sarge. here is the relevant part of mon.cf: watch testservers service http interval 15s monitor http.monitor period wd {Sun-Sat} alert test.alert -p 80 alertevery 45m upalert test.alert -p 80 alertevery 45m The alert script *is* being called, just not with the -u switch. Even if I add it in mon.cf _______________________________________________ mon mailing list mon@linux.kernel.org http://linux.kernel.org/mailman/listinfo/mon