3 Replies Latest reply on Feb 10, 2009 4:21 PM by wjcmalana

    Go red if EIGRP does not work

    Seashore

      Hi everybody

      A customer has a backupconnection. One time the primary connection died and the backup was to take over but it didn't work. The backupline didn't get any EIGRP-trafic. Polling hade worked befoure as the trafic took the primary line.

      The monitor that I guess I can make my own poller to monitor EIGRP on the backuprouter. I guess I can make an alert to alarm me if EIGRP does not work. (Please correct me if I'm wrong!!!) But can I make something go red on my map as well?

      (Or if so, is there a better way to solve this?)

      Regards

      Seashore

        • Re: Go red if EIGRP does not work
          Yann

          Hi,

          How would you define that EIGRP does not work? Is there a specific route you need to check?

          The following link details the EIGRP Mib and what can be checked:

          http://www.cisco.com/en/US/docs/ios/12_3t/12_3t14/feature/guide/gteigmib.html

          Once we know what to poll, we can easily setup a Trap Rule/Alert and/or a Universal Device Poller + an Advanced Alert to notify you.

          HTH,

          Yann

            • Re: Go red if EIGRP does not work
              Seashore

              Hi

               

              So far so good.

              But one question remains, can I get a map-object to go red with an advanced rule?

              If so I think I (or to be more correct my college) are happy.

               

              What happend once befoure was that the link was up but no EIGRP "notifications" (don't know the term) reached the router so it had no routes to use. You all know what happens then, no traffic.

               

              /seashore

                • Re: Go red if EIGRP does not work
                  wjcmalana

                  I have the same issue on my primary network on Qwest MOE.

                   

                  Our edge routers are connected to Qwest's 3750 Metro switches. The problem arises when the 3750 Metro loses layer 1 connection to the CO. The edge router is still reachable thru the backup network but it will not show any discrepancy since the physical interface between the edge rtr and the metro is still active (albeit no traffic).

                   

                  Not the cleanest way of doing it but i've setup interface traffic alert that get triggered when traffic on the interface is below certain figure. Also I setup up a Kiwi syslog to report EIGRP neighbor changes thru e-mail. Yes I get tons of email but i'd rather not have the Help Desk tell me that my network is down.

                   

                  Would also like to see some visual alerting thru the map as well. Can a 0 traffic on an interface trigger a node to go to "unknown" status?