>Okay, I have to vent here. This kind of question really bugs me. And not
>because of the answer, which I agree with. It just bugs me because the
>question assumes all routing protocols behave in the exact same way, when in
>fact they do not. My first thought when I read this question was "Hey - BGP
>prefers external over internal."
Well, I wrote the question, and it was associated with a paper on
IGPs. The challenge questions are a marketing technique -- questions
are extracted from the full question list for the currently
accessible free white paper, and aren't meant to be totally
free-standing.
The situation is more complex in exterior routing--note carefully I
am saying exterior routing, not purely BGP. Two paradigms apply to
provider routing, hot potato/closest exit and cold potato/best path &
exit. While it is true that the BGP route selection algorithm does
prefer eBGP to iBGP, real provider routing may include controlled
leaking of external information into internal, assuming a cold potato
model. In cold potato, you want to keep the path as internal as
possible, because that gives the greatest control, although it is
more resource intensive. Hot potato is less resource intensive but is
associated with far more variable performance.
I think you'll find this distinction drawn more carefully in
questions that were written to accompany the BGP white papers at
CertZone.
>
>So, question-writers, please take note: when writing a question, please
>write a *complete* question. In this case, if the author had simply
>metioned "IGP routing protocols," none of you would be suffering through
>this rant. ;-)
>
>We now return to our regularly scheduled mail list. Flame on.
>
>BJ
>
>
>
>----- Original Message -----
>From: Customer Service
>To: [EMAIL PROTECTED]
>Sent: Tuesday, March 06, 2001 10:56 AM
>Subject: This Week's Challenge Questions Posted
>
>
> ***********************************
>5) This Week's CCIE Challenge Question
> ***********************************
>What is the principal reason that external routes are
>always less preferred than internal routes?
>
>a) Performance is more predictable, since you know the
>end-to-end path for an internal route.
>
>b) Since you control all the resources used for an internal
>route, you have a better chance of troubleshooting it.
>
>c) It is a loop avoidance technique.
>
>d) Routers process internal route information faster than
>external information.
>
>The answer to this week's question can be found at:
>http://www.CertificationZone.com/QOW/1/TL/ccie-a.html
>
>
>_________________________________
>FAQ, list archives, and subscription info:
>http://www.groupstudy.com/list/cisco.html
>Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]