[193561] in North American Network Operators' Group
RE: BGP route processing speed
daemon@ATHENA.MIT.EDU (Sriram, Kotikalapudi (Fed))
Sat Feb 4 04:19:17 2017
X-Original-To: nanog@nanog.org
From: "Sriram, Kotikalapudi (Fed)" <kotikalapudi.sriram@nist.gov>
To: Sebastian Spies <s+Mailinglisten.nanog@sloc.de>
Date: Fri, 3 Feb 2017 17:00:04 +0000
In-Reply-To: <5890C175.9040307@sloc.de>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
>From: Sebastian Spies [mailto:s+Mailinglisten.nanog@sloc.de]=20
>
>my BSc thesis from 2010 might be relevant to what you are looking for.
>
>https://drive.google.com/file/d/0B5kLBHCcFJjFZk5RTUtwbUstbm8/view?usp=3Dsh=
aring
Thanks, Sebastian.=20
Good study. Your convergence time estimates (e.g. Table 2, p. 38) for route=
servers
are interesting and also consistent with the AMS-IX study in 2012=20
(the latter in realistic IXP scenarios).
But I am still interested in any pointers to=20
BGP router measurements in ISP/provider edge scenarios.
Sriram
>Sriram, Kotikalapudi (Fed) schrieb:
>> I am interested in measurements related to BGP route processing speed
>> (i.e. routing engine capacity in terms of routes or updates processed pe=
r second).
>> Folks from AMS-IX did an interesting study in 2012
>> in their Route Server / IXP environment.
>> https://ams-ix.net/downloads/ams-ix-route-server-implementations-perform=
ance.pdf=20
>>
>> Are there other measurement studies available
>> on this topic, especially in ISP/PE router scenarios,
>> including BGP policy processing, best path selection, route filtering, e=
tc.?
>> Will appreciate much if you can share some pointers.
>>
>> Sriram
