From: Howard C. Berkowitz (hcb@xxxxxxxxxxxx)
Date: Sun Jun 02 2002 - 09:04:45 GMT-3
>Howard,
>
>I agree with you completely. Perhaps I used the word *amuse* in a wrong way
>(sorry, if I have offended you/Nigel in any way by that).
Not a problem.
>I actually use BGP
>in my day to day work life, and certainly do understand the workings (of
>course, not at the level at which you would), but my sole purpose of finding
>an alternative way was *exam/lab* purposes only. Or rather should I say, I
>had a clue that I could use route-maps, but was lacking a *trigger/clue*
>
>I agree with you in that the scenario I put forward was absolute bizarre,
>bordering onto insanity. But prolly the same could be said about the lab
>(who said, they test us on production scenarios).
I tend to think of lab preparation strategy in what people variously
might call "force multiplier" or "statistical" approaches. There are
so many IOS knobs that no one has the time to go through every
possible combination while studying.
Bruce Caslow probably gave the best initial description of what I
believe to be the most effective learning model: focusing first on
what he calls the "graduate school" method of "issue spotting and
analysis." That's why I focus so much on "what is the problem you are
trying to solve," even in artificial scenarios based on complaints on
how insane they were. Once you can come up with a good problem
definition, you can ask yourself what sort of mechanisms would solve
that, and only then think of commands.
It's like "ready, aim, fire." Each step is important, but the
sequence is equally important.
For me, if I can't at least roughly describe the problem in even
pseudo-RPSL, that's a strong warning to me that I don't understand
the problem. This is especially true for BGP, but I personally use
the concepts for IGP problem that have area structures or some other
means of scoping the propagation of routing information.
>
>Once again, I wont let the opportunity go to mention that I really
>appreciate your understanding of subject matter and I really hold high
>regards for your knowledge & input.
>
>rgds
>Nick
This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:21 GMT-3