OSPF in core should converge in 5 seconds

 OSPF in core should converge in 5 seconds

 

As per Petr Lapukhov document on convergence,

here is sample configuration he provided in end of document:-

 

outer ospf 10
!
! Suppress transit link prefixes
!
prefix-suppression
!
! Wait at least 50ms between accepting the same LSA
!
timers lsa arrival 50
!
! Throttle LSA generation
!
timers throttle lsa all 10 100 1000
!
! Throttle SPF runs
!
timers throttle spf 10 100 1000
!
! Pace interface-level flooding
!
timers pacing flood 5
!
! Make retransmission timer > than arrival
!
timers pacing retransmission 60
!
! Enable incremental SPF
!
ispf


Please provide your views

Comments

  • I am not sure why you want to do it this way. Did you try other options like "Fast-hello", "BFD" and ISPF. I am sure those could have helped you to converge ospf under 5 seconds too [;)]. Make network type Point-to-Point and even you don't waste time in DR/BDR election [:)]

    But otherwise personally I have never seen someone playing around SPF and throttle timers in production. IMHO it's one of big collection of that useless stuff which CCIE teaches us.

     

    But who cares..but other wise you config looks good to me. But still only our CCIE Scientist Petr can confirm if it's 100 % correct. [:P]

    Deepak Arora



    http://deepakarora1984.blogspot.com/

  • you'll need to address many points to modify the overall convergence speed

    what is your exact interset here and what is your constraints ?

  • Wouldn't this be enough?

    interface X

       ip ospf hello-interval 1

       ip ospf dead-interval 5

     

  • that depends on the syntax of the question and what is intended by "convergence" here 

    is it just adjaceny control (esteblishment and failure detection) ?

    update advertising and propagation and processing and installation ?

    or both 

     

    i believe it is both , are you guys ?

     

Sign In or Register to comment.