HUM 1.0 Interface down

[Peter Valdes]

Hi,

Final demo, HUM. I'm currently testing alerts when an interface is on "down" status (unreachable) using a 3600 router (IOS C3620-J1S3-M ver 12.3(18)), to send email critical alerts. I have just added the device in the DCR and also added to Poller and Threshold. In the threshold setup, Im using variables ifOperStatus and the sysUpTime. On the trial on threshold, I disconnect the interface setup for the threshold and i'm not getting alerts or graphical visibility on the device availability HUM portel.

Main goal is to alert me via email when an interface is down.

My current setup in Threshold:

Template Name: Interface Availability

Variable Name: ifOperStatus

Condition: >=

Value: 2

No of Violations: 1

Sverity: Critical

Notify me

Any suggestions?

[Joe Clarke]

How long did you wait after pulling the cable? You will need to wait at most as long as your polling interval (which is five minutes by default).

[Peter Valdes]

Hi,

I waited almost 5 mins and given up. The Poll Interval is set to 1min.

Does this variable looks at the show ip interface brief "protocol or status"?

Some IOS has a bug that keeps the interface status on UP when removing the physical cable.

[Joe Clarke]

HUM looks at the ifOperStatus (SNMP object) of the specified interface. If the ifOperStatus is 2 (meaning down), then HUM will know the interface is down. You can check ifOperStatus by doing an snmpget on ifOperStatus for the given interface's ifIndex.

[Peter Valdes]

Hi jclarke,

How do I go about doing this task?

Is there another way of getting the LMS server to notify via email when the WAN interface goes down? It seems to not work and when it does, its 20 mins laters.

For example:

CiscoWorks HUM: Critical: 1 Threshold Violation: Test Down (10.0.0.69)

10.0.0.69 has violated the 3.0 1 threshold

Details:

MIB object name: locIfLineProt

Instance Name: 1

Number of violatons: 1

Threshold condition: var <= 3.0

Violation value: 1.0

Severity: Critical

Date and Time: Oct 16, 2008 3:42:00 PM

[Joe Clarke · Accepted solution]

HUM is not a fault management application. If you need better reporting on when a WAN interface goes down, considering configuring DFM to send email notifications on OperationallyDown events. Or configure RME's syslog automated actions feature to send you an email when it receives a syslog message indicating the WAN interface has gone down.