In <FDB040C62C7D26CA3B1E09CC at ctbeers-laptop.local> "Christopher T. Beers" <ctbeers at syr.edu> writes:
--On January 22, 2005 11:03:41 PM +0100 Henrik Stoerner <henrik at hswn.dk> wrote:
| Alerts in the current betas are rather flaky, unless you happen to | set things up just the way I do.
which is?? Would love to see a sanitized alerts cfg file.
It's not so much a question of how to setup the hobbit-alerts.cfg file, but more of doing things in the right sequence; since the current betas only trigger alerts when a status changes, you need the alert configuration to be setup and hobbitd running before the status goes red.
But since you asked, here's the alert config I use for the sites shown on http://www.hswn.dk/hobbit/
Two routers I was monitoring for just ping.
HOST=router1.aab11.dk,router2.aab11.dk SERVICE=conn MAIL henrik-aab11 at hswn.dk REPEAT=60
This is actually the most important rule to me. All hosts
except the listed ones match this rule, and I get alerts
for all services except "dnsinfo". red and yellow alerts
repeat every 15 minutes, purple ones only once a day.
HOST=!pto.linuxbog.dk,!router1.aab11.dk,!router2.aab11.dk SERVICE=!dnsinfo MAIL henrik at hswn.dk REPEAT=15 RECOVERED COLORS=!purple MAIL henrik at hswn.dk REPEAT=1440 COLORS=purple
And a simple rule for testing alerts via a script.
SERVICE=test SCRIPT /home/henrik/bbalert.sh henrik at hswn.dk RECOVERED
Henrik