From: Darby Weaver (darbyweaver@yahoo.com)
Date: Fri Dec 14 2007 - 03:18:16 ART
Cielieska,
That is a profound question and one very much
fundamental to all of us seeking this certification,
and one probably answered very well by those who have
already passed, and answered even better by those who
are instructors, of which at least two have already
supplied answers.
The trick is to understand L1 and L2, then be able to
configure them with both speed and accuracy.
(Personally, I think I pretty much got this part after
a while, but you gotta know, I'm a little slower than
the average slug it seems... but I get there and I
leave a trail that is easy to follow).
Now, one those areas are yours, and you say they are
so we'll move on.
Layer 3, this is a relatively straight forward layer
of the onion. Well, so it seems. However, I for one
have had some difficulty it seems here and since this
is the crux of you question allow me to elaborate just
a bit:
1. You absolutely have to own subnetting. I know
everyone thinks they do already. For many off us, we
are surprised at how easily we are mistaken on this
most fundamental of tasks to acquire. I mean 2 to the
2nd power minus 2 right? Well not quite.
There's a lot here in this most basic of L3 tasks.
You might need to understand summarization, creating
aggregates (notice I made a distinction), you will
want to understand prefix-lists, and access-lists with
wildcards as well. I mean overall there are a lot of
tasks here but what you really need to know is if
someone gives you a task like one of the following,
how fast can you complete each task and are you 100%
correct, because in a graded lab, each such task may
be worth 1, 2, or 3 points and more if other tasks are
somehow dependent on this task.
- All evens or All odds
- All evens within a given range
- All odds within a given range
- Filter by a given prefix, and a certain bit
boundary, or subnet, or maybe you will be asked to
permit or deny all specific addresses.
2. Now one you've classified your traffic, will you
use an acl, distance, a summary or range, a
distribute-list, a route-map, perhaps an aggregate
addition to BGP, maybe I've missed one in my haste...
3. Now are you going to add any conditions to whatever
you classified? Perhaps a metric? Perhaps another
condition and in which direction? What are you trying
to achieve? This is where it starts to get trickier,
right? I'd just call these modifiers.
4. But wait, there is more... what about other
protocols, and redistribution perhaps, or worse stub
areas of one type or another? And not that you are in
a spider's web of choices and decisions, perhaps there
is a bit of authentication as well (small matter but
to the un-initiated it can be troublesome and might
make one even remove a perfectly good authentication
configuration if one is unsure how to verify such
trite matters).
5. I did hear you say that Layer 1 did not bother you
as much, so much that you skipped it... Well, I'd say
if you skipped a thorough review of humble Layer 1 and
the attributes of a given interface, you might find
repeating the lab to be one of the things in your
future. You see... if you forgot about a little guy
called split-horizon, you may not get routes or
prefixes are needed somewhere else in your topology.
Now you would lose the 2-3 points there, maybe some
redistribution points, and for sure some filtering
points on one or both sides or more (maybe 3+) sides
of the equation. To make this stick watch the math:
3 Points for the interface configuration
3 Points for the redistribution (multiply X times the
number of protocols or tasks required)
3 Points for each filtering task.
==================================
9 Points minimum and maybe you might lose them and
their modifiers all over a little chum called
split-horizon. Even IPv6 may require it.
Bruce Caslow wrote a well-referenced book called
Bridges, Routers, and Switches and it speaks of a
layered approach to the lab. He starts with layer 1
and there are many good reasons to do so. If you do
not own a copy, you ought to let me know, I think I
own at least 3-4 copies of the second edition and at
least 2 copies of the 1st edition.
Now you might get the impression I think that book is
a cornerstone to one's CCIE studies. And if you did,
then you are right. However, some people may think
cause it covers things like ISDN, ATM, and DLSW that
it is out of date and un-important in todays one-day
lab.
Hmmm... is that right? Well, I still make it a point
to read that book from time to time. Why? Because if
I didn't I might believe I could only use either
frame-relay interface dlci or frame-relay map
exclusively, when in fact I can use both if I am
careful and if I might need to. But then you'd
probably ask why I might need to, and I'd have to tell
you to go get your copy of the book.
Also, it might sound strange, but Bruce and Val still
give this book away to all students of their NMC-1 and
NMC-2 classes, or at least as of last year.
Yep... Still think it is dated?
I've heard reference to bootcamps and there are
several good ones and both guys, Scott and Narbik have
one and I understand both are very impressive.
I've met Scott in-person now and I can see from his
verbal skills, mental tact, and body language that he
can definitely deliver the goods, and I've read about
a million of his helpful posts here on GS to boot. I
even learned a lot of what I know about frame-relay
from Scott through his VOD. He was even probably the
first guy to really tie down the need for a product
that covered the lab topic by topic in a workbook. I
have not yet been to the IPExpert Bootcamp - no
particular reason, only time it seems and convenience
have been at issue. Not much reason otherwise.
Recall Caslow and Satterlee also made a pretty good
name for themselves with their timeless books, both of
which are still useful in one's studies.
Now, I've seen a video from Narbik giving a lecture on
OSPF and we all thought it was most outstanding. I've
also had the pleasure of using a few of his topical
labs on specific subjects and found them both
thoughtful and useful. I have not been to his
bootcamp as of yet, but maybe this might be the year
to see what I have been missing.
Finally, if you really want to nail a topic like
routing and the intricacies involved, then you might
have to do what I've tried to do and maybe more of it:
NMC has a product called CheckIT. This product is the
most amazing little creation (thanks to Val aka Mr.
CheckIT). What it does is simple in scope but very
valuable in practice, let me elaborate a bit.
It allows you to configure your lab and then compare
it to a baseline configuration of a properly completed
lab.
Now, that does not sound much different than all of
the other workbooks does it.
But it is different and no one else has it or anything
like it yet.
You see we have COD's for labs, Jeremy created the CBT
Nuggets to help us out. The Brian's created the COD
to help us even more. And wonderful tools they are.
However, the CheckIT allows you and me to compare our
actual configuration and most show commands to the
correct configuration line by line.
Now with NMC-1 Labs, and Exercises, and NMC-2 Labs and
any graded CheckIT Lab, you can compare how you
actually configured your lab to the baseline and
perform show commands as well.
Now if you do not see a value in this product...
Really I mean, I've made it the Noon Feast in
NeverNeverLand too many times because I did not use
this tool (Ok, I'll admit it, I spent quite a while
being beaten up and intimidated by the level of
sophistication and complexity in the CheckIT Labs).
But, I'm now thinking I can perform most tasks
decently and quickly enough to engage this CCIE-Making
Tool once again and see where my own skills are still
lacking.
Trust me, I've still got some holes in my armor.
So... for me over the next few weeks, I'm revisiting
my own plan of attack aka my checklist.
I'll be taking Bob (NMC) Sinclair's advice and just
nailing up Multicast (not like I didn't do well enough
in graded labs enough, but in the lab, I need help so
I'm going back to my NMC Labs and others and am going
to own this topic).
Next, I'm going to give-in and buy Caslow's Video on
Cat-QoS - I sat the class and I know how to manipulate
DSCP and TOS bits, and yep, I can mark em and match em
thank to Bruce's explanation, but I'm not taking any
more chances on this topic. It has not really beat me
up yet, but again I'm once bitten twice shy now.
Let's see IPExpert's materials have some of the
coolest tricks I've seen yet and I really need to
invest a little capital in their December special.
I've learned a lot of time-saving tricks, but I'm
still "hearing from others" tips from Scott's lectures
and VODs, so it is probably about time I see for
myself.
If you haven't guessed, I've been humbled by the lab a
few times now. I've invested enough money to feed a
few families in certain areas of the world a few times
over in this quest and I'm nearly at the end of the
rainbow. My last two trips I've been very pleased
with my lab performance but still no shiny apple or
digits yet.
Obviously, I still need some work. So take my advice
with a grain of salt.
There's one more guy, who I think twists some of the
concepts you ask about to new and old heights and that
is "The Evil Bastard from Germany". He does a good
job of turning the page for a lot of CCIE's. He does
it with brute force and no mercy, he'll tell you those
weak need not apply. He wants those who are starving
for their CCIE and he takes them to the feast. The
"Evil Bastard" is brutal, since he gives you tasks
that the keyboard will not let you walk away from,
tasks that demand an answer. He's personal and makes
your experience more personal than mostly any other
can. His classes may be small and I've even heard of
one-on-one at times. He challenges your every move,
check and check-mate. He actually asks you questions
on a per task basis. When you are done you feel like
you've passed the lab is what some have said. Now,
I've not sat his 3-Week marathon class that is nothing
less than a CCIE-Factory (Let's face it, this guy only
does one thing -> He makes CCIE's); however, I've met
those who have sat the three week class and watched
some go straight to the lab and come out minted with
their own digits. This guy has given labs himself in
20 countries and on 4 continents.
Naturally if you are on the list, then you know the
Brians of InternetworkExpert. Out of all of the
vendors or trainers mentioned they have the most
CCIE's on their own Wall of Fame. They break down
topics, offer excellent tips and tricks and I've
learned a lot from them and am still learning a lot
from them. Their Class on Demand product, it's
low-price and it's quality is why a lot of CCIE's are
CCIE's today. Also both guys contribute a lot to this
list and other places as well.
NLI is CCBootcamp and they offer a range of courses
and a lot of tasks per workbook. The tasks are very
much in line with the tasks of the real labs, not like
some workbooks where tasks may be non-related or
overly complex to the point of being less than useless
except as tools for showing someone how unworthy they
might be. Brad Ellis is an owner of this fine company
and you will find some of the trainers and a lot of
the CCIE's on this list, of his list as well (The Hall
of Fame). That should say something. Also many
trainers have worked for these guys over the years as
well. They host a variety of products. I've used
their workbooks a bit over a version or two and found
their explanations very helpful.
And before the world rains in flame... I've given a
testament to a lot of vendors and a lot of products
above. Some of which I've used a little to a lot.
Some of which, I'm getting ready to schedule some labs
with to see if I've really grown with by comparison (I
have not attempted a graded lab now, in over a year).
Time to see how I stand in the eyes of the vendors who
helped shape and mold me.
And BTW - to go back to your question - you can use
each vendor's graded labs and or Cisco's own CCIE
Accessor Labs ($$$ but it is from Cisco). Remember
you can do 2 Cisco 1/2 labs for the same money you
might get 3-4 labs from a vendor or even a spinkle it
around and try a few vendor's graded labs.
Dollar for Dollar - I think this is probably one of
the best buys available in the market.
I also highly recommend you at least try each vendor's
free sample lab to get a feel for which vendor most
appeals to you and your own needs.
--- Cielieska Nathan <ncielieska@gmail.com> wrote:
> All Router and Switch gods,
>
> I have started my CCIE training about a month ago
> and i'm trying to
> tackle the topics as the OSI flows, for instance:
>
> Skipping Physical layer
> My first month would be Frame/Switching/DOT1Q
> Tunneling
> My second would be IPV6/Advanced IPv4 and IPv4
> routing .. etc etc
>
> So i'm into my first month and i have taken all IE
> Frame and
> Switching advanced labs and have bookmarked and
> highlighted overview
> documents on the subjects in question. Now i'm at a
> loss. I have been
> doing Spanning-Tree/Frame for years but knowing the
> "one-off"
> technologies (Storm Control) or Job Specific
> technologies (Like Dot1q
> tunneling would only probably be seen in decent side
> carrier/service
> providers) and knowing the material to the level the
> lab suggests
> leaves me in need for a methodical way to approach a
> topic, be
> proficient enough to pass with lab and move on to a
> new technology.
>
> My question is. Once you think you have a good
> fundamental knowledge
> of a topic what do you use to drill down and get to
> the intricate
> level that you need for the exam? And more
> important, when do you
> know when to move to the next topic. Any process or
> study tips would
> be much appreciated.
>
> Regards,
> Nate
>
>
This archive was generated by hypermail 2.1.4 : Tue Jan 01 2008 - 12:04:30 ARST