If the threshold of warning less than critical, snips thinks
the thresholds are increasing like 1 2 3. Othewise, snips thinks they
are decreasing like 3 2 1.
If the value of warning equals to critical, snips thinks they
are decreasing. So the thresholds of 2 2 2 means decreasing. And when
the value was 1, the state turns to CRITICAL.
Attached unofficial patch fixes this as increasing when the
warning equals to critical like nocol does. I think it's better for
compatibility. But it must be configurable.
md> I have a problem with snmpgeneric in snips-1.1:
md>
md> While this config works fine:
md>
md> TEMP 192.168.1.2 .1.3.6.1.4.1.2606.1.1.1.0 TEMP public 26 28 30
md>
md> This one doesn´t work:
md>
md> PRI 192.168.1.3 .1.3.6.1.2.1.2.2.1.8.2 STATUS public 2 2 2
md>
md> It seems that snips can´t handle the case when the thresholds for
md> warning, error and critical are equal although otherwise indicated
md> in the sample-config. The output goes to "critical" for any
md> snmpget-output, even if it is just "1".
md>
md> What I want to reach is to have an OK for output "1" and a
md> CRITICAL for anything higher.
md>
md> --
md> Gruss / Best regards | LF.net GmbH | fon +49 711 90074-27
md> Michael Dipper | Ruppmannstr. 27 | fax +49 711 90074-33
md> md at LF net | D-70565 Stuttgart | http://www.lf.net
--
Japan Information Processing Service,Co,Ltd.
2-4-24 Toyo, Koto-ku ,Tokyo, 135, Japan
marya at st jip.co.jp
--- lib/event_utils.c.orig Thu Aug 2 08:28:15 2001
+++ lib/event_utils.c Wed Jan 9 09:08:25 2002
@@ -158,7 +158,7 @@
int status = 1; /* assume UP */
if (crit_is_hi == -1) { /* set automatically */
- if (critt > warnt) crit_is_hi = 1; /* higher value is critical */
+ if (critt >= warnt) crit_is_hi = 1; /* higher value is critical */
else crit_is_hi = 0; /* lower value is critical */
}
|