>Ok.  I'm even more confused now.  So you guys are saying that IBGP 
>peers will never
>progragated its route to other IBGP peers by "no synchronization" - 
>if no IGP is
>running, except by Route Reflectors??  So what's "no 
>synchronization" used for?

I don't understand what you are asking. Route reflectors run iBGP.

The purpose of no synchronization is principally to speed convergence 
in non-transit AS.  As others have been said, its importance is 
vastly overstated.

>
>I have one more question:  Is it true that routes injected into BGP 
>within an AS carry a
>next hop attribute of the BGP router that first advertised the 
>route? Please explain.


That's the usual behavior, although an alternate next hop can be set 
when needed.  For purposes of the CCIE, I wouldn't worry about that 
case.

>
>Regards,
>Hunt Lee
>
>
>Howard C. Berkowitz wrote:
>
>>  >No worries John.  It was I who mentioned the devious nature of
>>  >classless and synch as well :)
>>
>>  Always remember that the best ISPs have no class.
>>
>>  >
>>  >Keep in mind that synch was designed for transit networks that have
>>  >transit providing routers which do not run BGP.  Back when the
>>  >internet was smaller I expect some designs had the IGP in an AS
>>  >carry the full table, or parts of it and hence it was relevant to
>>  >make sure your BGP and IGP were synchronized to ensure you didn't
>>  >blackhole routes.
>>
>>  Precisely. I don't have the document number in front of me, but the
>>  old RFC on BGP/OSPF interaction, which assumed this model, has been
>>  recategorized as Historic (i.e., nobody does this, don't try it, it
>>  was a blind alley)
>>
>>  >Today, BGP is run fully meshed with all transit providing routers in
>>  >an AS peering with IBGP and hence synch is a complete non issue.
>>
>>  Full mesh, of course, has its scalability issues, and we deal with
>>  iBGP scalability measures such as route reflectors.  There is a trend
>>  to have the main BGP at the edge, and to have principally an IGP in
>>  the provider core.  The core is stupid, and is traversed by MPLS
>>  tunnels -- the role of the IGP is to establish reachability for these
>>  LSPs, which run between BGP speakers on the edges.
_________________________________
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