- A virtual community of network engineers
 Home  BookStore  StudyNotes  Links  Archives  StudyRooms  HelpWanted  Discounts  Login
Re: Scenario Design: Comments Invited [7:41955] posted 04/21/2002
[Chronological Index] [Thread Index] [Top] [Date Prev][Date Next] [Thread Prev][Thread Next]

I'm not sure that I've ever seen anyone succeed at scenario type-2 without
spending quality time with packet captures (this is, of course, a
non-trivially different population that the set of individuals who merely
BELIEVE that they have succeeded at scenario type-20).

A lab presents a great opportunity for this type of study, because you can
see what some might consider to be low-level (though not as low level as,
say, the presence or absence of electrical current) consequences of
configuration changes, reinforcing any theoretical knowledge you might have
gleaned beforehand.

Scenario type-1 sounds like it might be amenable to an approach involving
automated enumeration, or some very mechanical manual procedure, at leat for
the less abstract topics or at the very beginning of study.

----- Original Message -----
From: "Howard C. Berkowitz" 
Sent: Friday, April 19, 2002 11:21 AM
Subject: Scenario Design: Comments Invited [7:41955]

> I'd like to start a discussion on the design of two kinds of scenarios:
>       1. lab preparation.  (problem recognition, speed building,
>          interaction among many protocols, time pressure, etc.)
>       2. In-depth understanding of protocols (seeing the effects of
>          alternative configurations, learning how to solve specific
>          problems with specific technologies).  Pure tutorials on
>          technologies complement these hands-on experiences.
> The two requirements, of course, are not mutually exclusive. #3 are
> scenarios that either statically or dynamically switch between the
> modes.
> It is my hope that this will stimulate community discussion involving
> both people who use scenarios and people who write them.
> Now, a disclaimer:  I work for Gettlabs and Gett Communications, the
> former of which runs a virtual rack service.  Gettlabs itself uses an
> open-source model for its own scenarios, as does Fatkid and some
> others. Gettlabs has partnerships with IPexpert and
> CertificationZone, which sell scenarios and supplemental materials.
> My comments here are intended to be neutral, and I will listen, learn
> and share with competitors.  I have discussed my intentions with Paul
> Borghese, and one of our agreements is that this is eligible to stay
> off the commercial list as long as I make free scenarios available.
> 1.  Lab Preparation
> -------------------
> Above all, these have to prepare you for pressure and ambiguity.
> A fairly basic question:  should all lab preparation scenarios be of
> 8-plus hour length, or two four-hour segments (forcing the disruption
> of a lunch break)?  Alternatively, is it acceptable to have sets of
> sub-scenarios that build on one another, so you can practice for an
> amount of time you have available, then pick up later on?
> I think it's a given that all you should be given is the addressing,
> etc., in the one day lab, plus instructions on what you should do,
> restrictions (e.g., no statics), and some criteria for judging
> success.  Estimated completion times/points also are important.
> An interesting question, however, is whether the scenario should
> include some of the sorts of things where it is fair (based on
> non-NDA statements of Cisco policy and the variations in proctors) to
> ask a proctor a question.  Should such points include things where
> variously the proctor will and will not answer, or even, in marginal
> cases, flip a software coin to see if the proctor will answer)?
> I believe it's realistic to be able to see a solved configuration,
> but, when you see it, you either should have demonstrated successful
> operation or accepted that you will accept losing points to be able
> to go on.
> I do not think that hints are appropriate in a lab preparation
> scenario, with the caveat that this sort of thing is quite
> appropriate to technology learning, and, as I suggested in #3 above,
> scenarios could be developed (possibly with a specific execution
> engine) that let you switch between preparation and learning modes,
> and even back.
> 2.  Technology Learning
> -----------------------
> My general approach to designing such things is again to start with
> instructions, initialization, etc., but to break the exercise into
> relatively small steps.  Each step will have hints available, and
> will be fairly small so you can look at the successive changes to the
> configuration that move you closer to your goal.
> One difference comes with the physical presentation of the scenario.
> If it is a printed document, should the hints be in-line with the
> text, or in a separate section so you will use them only if needed?
> If the latter, should they be on separate pages or at least have
> significant "spoiler space" between them so you don't inadvertently
> get an unfair clue to what is coming next?
> If the scenario is running interactively, should hints and hint
> answers only be available with a specific user action (clicking a
> link, opening a file, etc.)?
> What backup materials should be available for technology learning
> scenarios?  Is a bibliography necessary, and is it adequate?  Should
> there be actual tutorials available?
> Should learning scenarios routinely contain show command outputs as
> well as solved configurations, or should they simply suggest which
> show commands to use and what to look for in their output?  There
> will always be, of course, specific cases where the full display is
> appropriate.
> --------- semi-commercial but free content follows ----
> First examples:
>     There are several beta-version downloadable scenarios, which
> contain some interactive links, at the site. I am
> not completely happy with the display formats, and these will change.
> The only conditions for their use are:
>      1. They are copyrighted, but carry an automatic license for personal
>         use by the person downloading.
>      2. They may not be used commercially without Gettlabs written
> permission.
>         This includes both classroom and distance learning/virtual rack.
>      3. We ask that you do not send copies to others, but that each person
>         download their own copy. The simple reason for this is that the
>         scenarios are in frequent update and we want to be sure people get
>         the most recent version.
>     You are not required to run these on our racks, but, of course,
> we'd like you to. Some scenarios may depend on traffic generators and
> such which are not part of the scenario, but of the overall execution
> environment.
> Second examples:
>     I am actively putting together an FTP server that will have more
> scenarios, but initially will not be in pretty format but in lots of
> separate files.  While we experiment with display formats, this
> allows me to keep hints, solved configurations, etc., separate.  This
> server should start being available early next week.
>     This server will also have downloadable copies of lots of
> presentations of mine from NANOG, the IETF and IRTF, ARIN, etc., as
> well as other recommended reading.  There will be some subdirectories
> labeled "working" that contain documents actively being worked on by
> teams/committees, and these may not make sense to anyone other than
> the coauthors.
>     Some of these presentations may be a little old, and I'll be updating
> them.
>     Warning, with half a smiley:  my ISIS tutorial may carry a curse.
> I tried to present it at NANOG twice. The first time, I came down
> with a flu bug that had me down for a good six weeks.  The second
> time, I had to have a cardiac pacemaker installed the day it was to
> have been presented. You Have Been Warned. There May Be Things That
> Man Is Not Meant To Read. (or, as a bumper sticker some will
> recognize says, "Vote for Cthulhu. Why settle for the lesser of two
> evils?")
> --
> "What Problem are you trying to solve?"
> ***send Cisco questions to the list, so all can benefit -- not
> directly to me***
> Howard C. Berkowitz      hcb@xxxxxxxxxxxx
> Chief Technology Officer, GettLab/Gett Communications
> Technical Director,
> "retired" Certified Cisco Systems Instructor (CID) #93005

Message Posted at:
FAQ, list archives, and subscription info:
Report misconduct and Nondisclosure violations to abuse@xxxxxxxxxxxxxx