How can i generate the AS path from IRR data same as found in BGP

Randy Bush randy at psg.com
Fri Jan 1 10:54:31 UTC 2010


>>> How similar do you think the paths would be from BGP and IRR data?
>> they are both ordered non-null sets of positive numbers drawn from the
>> range 1..32k
> so that means they can't be the same..
>    Due to IRR consistency issues ?? or why?

the irr was sabatoged early by some net-egos who did not think it was a
good idea.  major providers never registered their information.

it is pretty good in japan (i.e. if all ASs in the path are in japan),
mediocre in some of europe, and generally really bad elsewhere.

fwiw, this saddens me deeply.  i had pushed the irr for almost 15 years
before giving up.

> I have reviewed some of the previous work like Nemecis, RIPE RRCC,
> Univ.of Roma Analysis of RIR. What i have found is most of the work is
> on evaluating the consistency of IRR's.

american idiom: where there is smoke there is fire.

i.e. if the irr data were good, no one would bother analysing it.

> There are some other works on using IRR along with BGP and
> traceroute's to generate the topology of internet.

a very misguided exercise.  see our imc paper
<http://portal.acm.org/citation.cfm?id=1644893.1644923&coll=portal&dl=ACM&type=series&idx=SERIES10693&part=series&WantType=Proceedings&title=IMC>

> AS i am first year phd student at Seoul National University,Korea and
> working on BGP Security. For that i want to evaluate how i can use IRR
> data? 

another american idiom: you can't make a silk purse out of a sow's ear.

<http://archive.psg.com/091007.nag-rpki+origin.pdf>

randy



More information about the irrtoolset mailing list