From: kris.keen@xxxxxxxxxx
Date: Mon Aug 12 2002 - 22:07:53 GMT-3
Yep, its was poor reading on my behalf only. I understand fully now,
Thanks.
Onward with the reading!
-----------------------------------------------------------------
Kris Keen - CCNP, CCDP, CNE
Network Support Specialist - Network Systems
Aon Risk Services Australia Limited
(612) 9253 7272
0404862970
E: Kris.Keen@aon.com.au
"Howard C.
Berkowitz" To: kris.keen@aon.com.au, ccielab@
groupstudy.com
<hcb@gettcomm. cc:
com> bcc:
Subject: Re: BGP MED IRA Page 167
13/08/2002
11:03 AM
At 8:41 AM +1000 8/13/02, kris.keen@aon.com.au wrote:
>Hi All,
>
>Reading MED this morning, there is a diagram o page 167, Figure 6-13. It
>explains that a MED with a lower value takes perference, but reading the
>explaination of the figure it says it will prefer the SJ router (MED=120),
>I would have thought the NY router with MED=50 would have been preferred?
>
>It says it is comapring 3 different sources.. Is it just to early in the
>morning or am I clearly missing something?
>
>Cheers
>
always-compare-med will do this.
A real-world application is at a multilateral exchange point, where
you peer with several providers. There's an informal convention of
advertising MED as delay-based, so you can pick the "best" route.
Avi Freedman discusses this in several NANOG presentations -- Denver,
I'm pretty sure, was one, where I did the intro and he did the
advanced, hoping the tear gas from the street didn't get up to the
lecture room. Bronco fans were VERY happy.
This archive was generated by hypermail 2.1.4 : Sat Sep 07 2002 - 19:48:23 GMT-3