From: Peter van Oene (pvo@xxxxxxxxxxxx)
Date: Mon Feb 04 2002 - 17:59:00 GMT-3
It is not unheard of for course material to be incorrect :) As far as
choosing between primary and secondary source material, I tend to lean on
the former.
At 07:10 PM 2/4/2002 +0000, Mark Lewis wrote:
>As far as iBGP behaviour not constituting 'split-horizon', Cisco
>themselves refer to the previously referenced behaviour as 'BGP
>Split-Horizon' (see the BSCN & BSCI courses). However, the RFCs do not, as
>far as I'm aware, refer to it as 'split-horizon', so take your pick.
>
>Rgds,
>
>Mark
>
>
>>From: Peter van Oene <pvo@usermail.com>
>>Reply-To: Peter van Oene <pvo@usermail.com>
>>To: John Neiberger <neiby@ureach.com>, "Mark
>>Lewis" <markl11@hotmail.com>, ccielab@groupstudy.com
>>Subject: Re: Re: Junos BGP routes update
>>Date: Mon, 04 Feb 2002 12:05:53 -0500
>>
>>Issue is likely even simpler and the question stems likely from a variance
>>in behavior. If A EBGP peers with B, in Juniper, A will send directly back
>>the B the paths received from B. B then drops per the AS-Path
>>loop. However, Cisco's implementation will automatically prune these
>>advertisements in a split horizon like method.
>>
>>Also, as previously discussed (though maybe on the juniper list?), IBGP
>>advertisement rules do not constitute a split horizon implementation.
>>
>>At 10:57 AM 2/4/2002 -0500, John Neiberger wrote:
>>>That statement is certainly true but I thought we were
>>>discussing eBGP, not iBGP. I thought he was referring to a
>>>scenario where an eBGP peer receives an update from another
>>>eBGP peer that includes its own ASN.
>>>
>>>Let's say AS1 has routers A and B, both of which are connected
>>>to AS2, routers C and D. Both A and B advertise their prefices
>>>to C and D who in turn trade updates. Then C will advertise
>>>back to AS1 the routes it learned from D. The receiving router
>>>in AS1 will see its own ASN in the AS-PATH and drop the update.
>>>
>>>I believe that the original poster thought that AS2 wouldn't
>>>even transmit an update to AS1 that had AS1 in the AS-PATH.
>>>I'm fairly sure that it is up to the receiving router to check
>>>the AS-PATH for its own ASN, not the advertising router.
>>>
>>>Regards,
>>>John
>>>
>>>
This archive was generated by hypermail 2.1.4 : Thu Jun 20 2002 - 13:46:11 GMT-3