nanog mailing list archives
Re: Long AS Path
From: Pierfrancesco Caci <pf () tippete net>
Date: Thu, 22 Jun 2017 08:09:41 +0200
"Mel" == Mel Beckman <mel () beckman org> writes:
Mel> Why not ask the operator why they are pretending this path? Perhaps
Mel> they have a good explanation that you haven't thought of. Blindly
Mel> limiting otherwise legal path lengths is not a defensible practice, in
Mel> my opinion.
Mel> -mel beckman
A prepend like that is usually the result of someone using the IOS
syntax on a XR or Junos router.
Long ago, someone accidentally prepending 255 times hit a bug (or was it
a too strict bgp implementation? I don't remember) resulting in several
networks across the globe dropping neighbors. One has to protect against
these things somehow.
As a data point, here is how many prefixes I see on my network for each
as-path length, after removing prepends:
aspath length count
-------------------------
0: 340
1: 47522
2: 292879
3: 227822
4: 58390
5: 10217
6: 2123
7: 638
8: 48
9: 58
11: 20
12: 2
So, does your customer have a legitimate reason to prepend more than 5
times? Maybe. I still think that anyone that does should have their BGP
driving licence revoked, though.
Pf
--
Pierfrancesco Caci, ik5pvx
Current thread:
- Long AS Path James Braunegg (Jun 20)
- Re: Long AS Path Olivier Benghozi (Jun 20)
- Re: Long AS Path Laszlo Hanyecz (Jun 20)
- Re: Long AS Path sthaug (Jun 21)
- Re: Long AS Path Randy Bush (Jun 21)
- Re: Long AS Path Stephen Satchell (Jun 21)
- Re: Long AS Path sthaug (Jun 21)
- Re: Long AS Path Mel Beckman (Jun 21)
- Re: Long AS Path Pierfrancesco Caci (Jun 21)
- Re: Long AS Path jim deleskie (Jun 22)
- Re: Long AS Path Mel Beckman (Jun 22)
- Re: Long AS Path Bob Evans (Jun 21)
- Re: Long AS Path Saku Ytti (Jun 21)
- Re: Long AS Path Jon Lewis (Jun 22)
- Re: Long AS Path Stephen Satchell (Jun 22)
- Re: Long AS Path Ryan L (Jun 23)
