[c-nsp] SNMP OID to see if a Tn interface is looped up?

[Ryan West]

Good find.  I was pretty sure that the operstatus MIB had more than 2 values.  We use something similar with Zabbix.  FYI, here are the integer values:

ifOperStatus OBJECT-TYPE
    SYNTAX  INTEGER {
                up(1),        -- ready to pass packets
                down(2),
                testing(3),   -- in some test mode
                unknown(4),   -- status can not be determined
                              -- for some reason.
                dormant(5),
                notPresent(6),    -- some component is missing
                lowerLayerDown(7) -- down due to state of
                                  -- lower-layer interface(s)
            }

-ryan

-----Original Message-----
From: Justin Shore [mailto:justin at justinshore.com] 
Sent: Thursday, July 16, 2009 7:20 PM
To: Ryan West
Cc: 'Cisco-nsp'
Subject: Re: [c-nsp] SNMP OID to see if a Tn interface is looped up?

Ryan West wrote:
> Justin,
> 
> Give this a shot:
> 
> http://tools.cisco.com/Support/SNMP/do/BrowseMIB.do?local=en&step=2&mibName=CISCO-ICSUDSU-MIB
> 
> That MIB contains values for different loop codes.

Ryan,

That looks like a very useful MIB.  I'll give that a try.

I looked at the source of the check_snmp_int.pl script for Nagios as 
well and noticed that it was built to handle 3 response codes:  up, down 
and testing.  On a hunch I looped up an unused T1 on a T3 controller and 
hit it with a snmpwalk.  Sure enough it listed the looped interface as 
"testing":

IF-MIB::ifOperStatus.28 = INTEGER: testing(3)

check_snmp_int.pl also correctly detected the situation:

Serial1/0/10:0:TESTING: 1 int NOK : CRITICAL

So it looks like that plugin will do what I need if I can figure out how 
to make it only give a warning if the interface is in testing and not a 
critical alarm (warning for say 1hr; then escalate it send our alerts). 
  I want it to warn with no email for about an hour.  Then escalate it 
and send our the alerts.  Should be doable.

Thanks
  Justin