2 Replies Latest reply: Apr 5, 2012 2:33 AM by shreeram RSS

    HSRP secondary keeps trying to become the primary

    Sidney
    Sidney Mar 21, 2012 10:50 PM

      Hello Guys,

       

      Our network consists of 2 core switches (6509-E) and 12 access switches. We have the core switches configured with HSRP for all vlans. 2 days ago we had to shut down the secondary core switch because of UPS maintainence. Since it was the standby it caused no downtime. After the reboot, i have noticed that the sh log command shows that the standby switch is going through the stages of speak->active->speak->standby

       

      This happens every few hours. I have checked the links between the 2 switchesand see no packet drops or errors. A sh log command on the primary does not show any logs related to the HSRP state changes.

       

      This is my HSRP config for one of the VLANS on the primary core:

       

      int vlan 2

      ip address 10.10.200.2 255.255.255.0

      standby version 2

      standby 1 ip 10.10.200.1

      standby 1 timers 1 3

      standby 1 priority 150

      standby 1 preempt delay minimum 180

       

      This is the config on the secondary core:

      ip address 10.10.200.3 255.255.255.0

      standby version 2

      standby 1 ip 10.10.200.1

      standby 1 timers 1 3

        standby 1 preempt delay minimum 180

       

      I have noticed that this happens around the time the cpu history shows a spike in utilization even if it is only 60% odd. Any thoughts on the matter would be appreciated.

       

      Regards,

      Sidney

         
        • 1. Re: HSRP secondary keeps trying to become the primary
          gloryitak
          gloryitak Apr 11, 2012 10:47 AM (in response to Sidney)

          Hi Sidney,

           

          How about you add "standby 1 preempt" to the primary core and see what happens.

           

          Message was edited by: Glory

          • 2. Re: HSRP secondary keeps trying to become the primary
            shreeram
            shreeram Apr 5, 2012 2:33 AM (in response to gloryitak)

            Hi Sidney & Glory,

             

            I think the problem is not with HSRP and its configuration. because,

             

            1. standby 1 preempt delay minimum 180 is The preempt delay feature allows preemption to be delayed for a configurable time period (in our case 180), allowing the router to populate its routing table before becoming the active router.

            so its ok

             

            but still I will suggest you to add

             

            1. priority on your stand by (defunately it will be less than 150 and it is best          practice).

            2. enable feature of preempt on both active and standby.

            3. enable feature of tracking on both active and standby (you should track your    WAN or next hop interfaces, again this for best practice).

            4.and instead of using timers try to use FHRP which enables feature of BFD in  HSRP (http://www.cisco.com/en/US/docs/ios/12_4t/12_4t11/hsrpbfd.html)

             

            Again again, what i suggested above are the best practice config tips.

             

            Now come to the point, As I told , I do not suspect with HSRP configuration.

             

            I am suspecting issue with one important thing you have mentioned above is that CPU utilisation. (But you have not mentioned router, standby or active)

            You have mentioned that the in time suddenly you are getting spikes in utiliston.

            First you have to fix this issue.

            Because, due to high utilisation your processor is not able to process packets and it is getting dropped and hence router is not able to send control packets and hence standby is not recieving control packets and attempting to become Active.

             

            To do this, you can use

            router#show processes

            CPU utilization for five seconds: 0%/0%; one minute: 0%; five minutes: 0%

             

            here importat part of output is CPU utilization for five seconds: 0%/0%

            left most % is your CPUs overall utilisation and rightone is your Data plane or forwarding plan utilisation

            By subtracting right one from left one you will get your control plan utilisation.

             

            Ideally, Control plan utilisation is very less say 15-25%

             

            But if you are getting high CPU utilisation you have to check why it is?

            If it is no because of control plane and due to forwarding plan that you have to check on which interface you are getting excesssinve traffic. (though this is lenghty process you have to do that).

             

             

            One more thing I would like to advise you that, always try to keep same image/version file on same .