4 Replies Latest reply on Mar 2, 2011 2:43 AM by Ciag

    Last status change incorrect date

    Ciag

      Hi

      I am using the switchport mapper for the first time since we acquired the Engineers toolkit. I have noticed when I include the 'Last Status Change' feild in the report a lot of the dates that are returned are wrong. I know this because a lot of them are dated 2010 and 2011.

      Is this a known issue with switchport mapper or is it a problem with counters rolling over or something?

      Cheers

      Ciaran

        • Re: Last status change incorrect date

          Hi Ciag - This is a known issue due to the way the ifLastChange date is presented to us in the counters.  It is the value of sysUptime when the change was made, so if sysUptime has rolled over, and becomes less that the last change counter value, it seems to us that the change happened in the future.

          We have been adding code to not show these dates in the future, and will make sure we have a tracking item to get this corrected in both SwitchPortMappers.

          HTH,

          Greg

            • Re: Last status change incorrect date
              Ciag

              Thanks for your reply Greg - so what you're saying is apart from continualy polling every device and storing the interface status change dates in a database there is no way of getting accurate dates for devices that have long uptimes? Is it being completely dropped? Would there not be any other counters used to determine the last int status change?

              Cheers

                • Re: Last status change incorrect date

                  Ciag - None that I am aware via SNMP. The most reliable and accurate method would be to CLI to the device and "show int".

                  HTH,

                  Greg

                    • Re: Last status change incorrect date
                      Ciag

                      Hi everybody,

                      does anyone know of any free tool that is able to accurately work out the last interface state change date regardless of sysuptime counter rollover?

                      Alternatively I have read that instead of using the sysuptime counter the SNMPengineTime (.1.3.6.1.6.3.10.2.1.3) counter can be used as it doesn't rollover for 135 years. This counter is in seconds and has been subject to bugs in certain versions of IOS where it did reset when sysUptime rolled over.  I have tried this on my 4500's and it looks to have reset with the last SysUpTime rollover.

                      I'm particularly interested in finding out an IOS image for a 4500 that is not affected by the snmpenginetime reset bug, or an image that uses a 64bit counter for SysUpTime instead of the 32bit counter that is limited 496 if such images exist I haven't been able to find out.

                      Regards