link error trap

[glen.grant]

Is there any way to trap on links , specifically uplinks that may be running dirty and taking errors on either catos boxes and IOS boxes . On the catos side we have boxes anywhere from 4.5.14 to 6.4.11 . The IOS side runs 2924's ( 4and 8 meg), 2950, 4506's , 3550's . Is there is anything we can implement that would tell us when the links are running less than desirable but are still in the up status ?

[Georg Pauwen]

Hello Glen,

you could try RMON (Remote Monitoring), which lets you create event driven SNMP traps based on any MIB value that a device supports.

Basically, you need to find the MIB value you want to monitor, and then create the alarm. Check this document for configuration details:

Configuring RMON Alarm and Event Settings from the Command Line Interface (CLI)

http://www.cisco.com/en/US/tech/tk961/technologies_tech_note09186a0080094478.shtml

HTH,

GP

[todd.martin-02]

Glen,

I would get this information from polling the device with you NMS. Trapping is a great way to gain visibility into a device status but is should be used in conjunction with Polling.

Polling the specicic device on a regular interval say 10 Min. will enable you to develop a baseline and then from there you can set up your RMON thresholds after you know what is "normal" performance on that link.

[wkumari]

What you are looking for is some sort of network management / monitoring application. Luckily there are a bunch of good free ones - OpenNMS (http://www.opennms.org) and Nagios (www.nagios.org) are two very popular. Both can easily be set to monitor (via SNMP) for and sent alerts via email / pager / SMS/anything else. You can also get them to generate SNMP traps to an existing NMS system if that strikes your fancy. Searching Google for SNMP error counters will heolp you find OIDs for usefull errors, or you could always download hte MIBs from Cisco.

Warren.