Inline

At 01:16 PM 9/10/2002 +0000, Karl Brusen wrote:
>Can anyone point me towards a resource that describes in detail what it
>takes for BGP to consider a route synchronized when it learned it through
>IGP OSPF?  My study partner and I are struggling understanding how it works.
>All of our resources provide only general information with statements like,
>"a route must be learned by IGP before BGP will consider it synchronized".
>Merely learning a route from IGP is apparently not good enough.  There must
>be other, more specific requirements.  How does route-reflection affect
>BGP/OSPF synchronization?

These technologies are not designed to work together.  Who knows if the BGP 
synch code even works. I fully expect the systest folks at Cisco who do 
regression testing have far better things to test (CEF comes to mind) than 
obsolete, fully deprecated behavior as used in broken network 
designs.  When building your "lab" network, use one of these scaling tools 
or the other.  In the real world, turn off synchronization.  If Cisco 
chooses to test scenarios of this nature then shame on them.  There are far 
more relevant design complexities that one should entertain oneself with.

For what its worth, the CCIE-LAB archives are riddled with discussion on 
this topic.

>Specifically, we are working on a lab with three routers running BGP.  They
>are in the same AS (IBGP).  They are not fully meshed one is acting as a
>router reflector for the other two.  One of the RR clients has a loopback
>injected into BGP by redistribute connected.  All three routers are also
>running OSPF and have an IGP route to the same network being injected into
>BGP.  The route is synchronized on the router it is injected into and on the
>route reflector, but it isn't synchronized on the other route reflector
>client.  We are aware of how BGP and OSPF router IDs can prevent
>synchronization, so we have specified the same router IDs for BGP and OSPF.
>
>What is interesting is that if we point a static route from the problem
>route-reflector client to the BGP route "next hop", BGP synchs.  Due to the
>network topology and modifications of the ad distance, the problem router
>also has a route learned from EIGRP but is not normally in the IP routing
>table.  When we shut down an interface so that the EIGRP route is placed in
>the IP routing table, BGP synchs.

>We are very confused.  There is something about how BGP synchs with OSPF
>that we just don't understand.  Any insight from the group will be greatly
>appreciated.
>
>Thanks,
>
>
>Karl Brusen




Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=52998&t=52980
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]

Reply via email to