[j-nsp] optimized switchover

[Matthias Gelbhardt]

Hello David,

great tip. Unfortunatly BFD for BGP - though detailed documented - has 
no examples flying around. Perhaps I am missing something here.

I have two routers connected via iBGP. I have tried to make the 
configuration rather simple (only the important parts, BGP session is up 
and running):

This is the same on both sides (change in the IP-addresses of course)

protocols bgp {
     group internal {
     type internal;
     neighbor 91.190.xxx.xxx {
         local-address 91.190.xxx.xxx;
         bfd-liveness-detection {
             minimum-interval 1000;
             multiplier 3;
         }
     }
}

Router A:
show bfd session extensive
                                                   Detect   Transmit
Address                  State     Interface      Time     Interval 
Multiplier
91.190.xxx.xxx           Init                     3.000     1.000 
  3
  Client BGP, TX interval 1.000, RX interval 1.000
  Session down time 00:00:04
  Local diagnostic CtlExpire, remote diagnostic None
  Remote state Down, version 1
  Min async interval 1.000, min slow interval 1.000
  Adaptive async TX interval 1.000, RX interval 1.000
  Local min TX interval 1.000, minimum RX interval 1.000, multiplier 3
  Remote min TX interval 1.000, min RX interval 1.000, multiplier 3
  Local discriminator 1, remote discriminator 1
  Echo mode disabled/inactive, no-absorb, no-refresh, update-adj
  Multi-hop, min-recv-TTL 0, route table 0, local-address 91.190.xxx.xxx

1 sessions, 1 clients
Cumulative transmit rate 1.0 pps, cumulative receive rate 1.0 pps

Router B:
show bfd session extensive
                                                   Detect   Transmit
Address                  State     Interface      Time     Interval 
Multiplier
91.190.xxx.xxx           Down                     0.000     1.000 
  3
  Client BGP, TX interval 1.000, RX interval 1.000
  Local diagnostic None, remote diagnostic None
  Remote state AdminDown, version 1
  Min async interval 1.000, min slow interval 1.000
  Adaptive async TX interval 1.000, RX interval 1.000
  Local min TX interval 1.000, minimum RX interval 1.000, multiplier 3
  Remote min TX interval 0.000, min RX interval 0.000, multiplier 0
  Local discriminator 1, remote discriminator 0
  Echo mode disabled/inactive, no-absorb, no-refresh
  Multi-hop route table 0, local-address 91.190.xxx.xxx

1 sessions, 1 clients
Cumulative transmit rate 1.0 pps, cumulative receive rate 0.0 pps

I see the diagnostic on router A but do not understand it. I thought the 
minimum-interval might be too low, so I set it up to a thousand.

Regards,

Matthias


David Ball schrieb:
>   There are likely several answers to that, all dependant on your
> topology and protocol use. But, a good place to start would be BFD
> (bidirectional forwarding detection).  Juniper has decent support for
> it working with other protocols (OSPF, ISIS, BGP, RIP), notifying them
> that something may be wrong, allowing them to then make a decision
> (support may differ from protocol to protocol).  That may be a good
> start point.
> 
> http://www.juniper.net/techpubs/software/junos/junos95/swconfig-routing/swconfig-routing-IX.html#B
> 
> David B
> 
> 
> 2009/9/6 Matthias Gelbhardt <matthias at commy.de>:
>> Hi!
>>
>> I wonder what the best practices for optimized switchovers would be? I mean
>> fast comprehension of failed BGP connections? A fibre cut or something like
>> that, how can I be sure, that my routers are detecting the failed session as
>> soon as possible? What would be the best practices fpr that?
>>
>> Regards,
>>
>> Matthias
>> _______________________________________________
>> juniper-nsp mailing list juniper-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/juniper-nsp
>>