17 Replies Latest reply on Aug 22, 2017 6:56 PM by superfly99

    Monitor CRC error  in cisco devices

    Max Tan

      Dear Friends


      im having trouble where i suppose to monitor crc error in Cisco devices with custom poller.

      i manage to create with CRC monitor with OID but unfortunately its unable to give me the correct CRC figure.

      The value in show interfaces showing 0 but from the polling result it showing 110255 which really making me confuse.

      Anyone can solve my problem?

        • Re: Monitor CRC error  in cisco devices
          Craig Norborg

          Just a guess here, but when you monitor something like that via SNMP, you get the count since the router was last rebooted.   If you look at a show interfaces, you get the count since the "clear counters" command was last issued.  Typically you would track CRC's over time and graph them so you can see when they happen as well as an idea of how much at different points in time.

            • Re: Monitor CRC error  in cisco devices
              Max Tan

              I got your point Craig, but how could i monitor crc count on real time basis?  because we want the alert triggered when the CRC count hit certain treshold.

                • Re: Monitor CRC error  in cisco devices
                  Craig Norborg

                  Well, I think your confusing alerting with monitoring.  Alerting when you hit a hard threshold on something like CRC's isn't too useful.


                  Here is why.  If you have two routers you enable this threshold alert on, both freshly rebooted so they both have a 0 count on the SNMP counters.  Let's say that router #1 hits the threshold, lets say 1000 CRC's within the first hour.   You will get the alert right then and there.   If you have it re-alerting due to the condition still being met after that, you will continue getting alerts.   However, if you only have it alert when it hits 1000 CRC's, you will get that one alert even if it is generating thousands of CRC's per-minute.


                  However, if router #2 gets 1 CRC error a day, it will take over 3 years to get your first alert from this router.  Wouldn't make sense to get an alert when you only see 1 CRC/day on an interface...


                  So, down to what you need to do, sorry to be a bit vague but while I know the process I don't have an example.   I believe you need to create your UnDP as a "Rate" rather than a "Counter" or "Raw Value". 


                  Then, in the Alert Manager, you have to set up an alert where the trigger condition is based on the "Type of Property to Monitor" being "Custom Interface Poller".   Then having the "Poller Name" under "Custom Interface Pollers" being equal to whatever name you assigned to your poller.   You should also have another condition where the "Custom Interface Pollers" /  "Rate" being greater than whatever value you want to alert on, whether it be 5 CRC's per second, or 100 CRC's per hour. 


                  I think you'll need to play with this a bit in both the UnDP, setting your polling rate and timeframe and such, and have it match with the Alerting too...


                  After that it should be pretty much a normal alert, set the action and reset conditions and such...


                  Hopefully this gets you on the right track!!

                  • Re: Monitor CRC error  in cisco devices

                    Craig has alluded to this but you could monitor on an hourly basis. So say if CRC's hit over 100 in an hour, then you can be alerted. That way it is counting on a per hour basis. So you'd need to decided what the high number is (go by previous experience) and then if the error count goes over that within an hour, you will be alerted straight away.

                    • Re: Monitor CRC error  in cisco devices

                      We do exactly as superfly99 mentioned over 1hour.

                      Anything else may not be user impacting as determined by our networks team nor the carrier want to truly investigate a low amount of errors over a 1hr period.

                        • Re: Monitor CRC error  in cisco devices
                          Max Tan



                          I did as what you all have suggested here but again please refer the attached.


                          The CRC value in physical server (use "show interfaces" command) and CRC value from Orion different and not tally how could this be happen?


                          please refer the attached image, Any Idea?


                          Interface value.pngOrion value.png

                            • Re: Monitor CRC error  in cisco devices

                              What's the device and IOS you're running as sometimes the OID is slightly different

                                • Re: Monitor CRC error  in cisco devices
                                  Max Tan

                                  Its cisco router modal 1841 and the OID used in Orion custom poller is locIfInCRC

                                  its keep return the same value eventough i have clear the counting in the server with "Clear counters" command.

                                    • Re: Monitor CRC error  in cisco devices

                                      We get the CRC errors out of the box, I don't believe there is a need to go down the path of setting custom UnDP to do that.

                                      We have the same setup of C1841 with a Serial Link running IOS 12.4(17). Details are straight from Solarwinds WebUI.
                                      If you do need to use the UnDP as well however then note sometimes the OID can change at times depending on the IOS.

                                      Are you actually incurring any errors since you started testing the UnDP you set up before clearing counters?

                                      Do you have a device you can properly test against, or a link to have known errors against?

                                      Maybe request a change window on this link if that's the only option you have.
                                      If you poll and get the same error count vs CLI showing errors then you need to check what your doing with the counter as well.


                                      Node Details:



                                      Operating System

                                      12.4(17), RELEASE SOFTWARE (fc1)
                                      IOS ImageC1841-ADVIPSERVICESK9-M


                                      Interface Details:

                                      NameSerial0/0/0 · ** Access Link **
                                      Alias** Access Link **
                                      Interface TypeSerial


                                  • Re: Monitor CRC error  in cisco devices
                                    Craig Norborg

                                    Look at your output where it says "Last clearing of "show interface" counters 3d14h.  As I said in my previous post, usually the only time SNMP counters (ie: what you poll via UnDP) and "show interface" counters match is when you haven't cleared the SNMP counters since the router rebooted.   You, or someone else, cleared the "show interface" counters 3d14h ago, so odds are your counters won't match.   What you are seeing is "normal" and correct behavior...


                                    Also, how is your UnDP set up?  As a counter, raw, rate?   This could influence how your values are shown on web pages, although in the screen capture above I believe you are looking at the raw counter variables from within the UnDP manager.