I passed my first attempt over a month ago at San Jose, #14023. I wasn't in
the mood to do a write-up until now...
I haven't posted to this group because I never felt had anything to say. But
I've been reading Groupstudy since the days of Token Ring switching, and to
all whose posts helped me, I can only say thanks.
Now, what you want to know... How was the lab? Overall, it was tough, but
I'd been using Internetwork Expert's labs, and I heard that the real thing is
a lot easier. To my suprise, I found it to be about an 8 in terms of IE's
labs. Maybe a 7. It was quite hard. That should should warn you to be very,
very prepared. When I first read the test, I saw a number of things I hadn't
seen before, as well as some questions where I initially wasn't sure what they
wanted me to do. This is why it is so, so important to know the theory. After
my initial panic, I dug in and figured out the answers pretty quickly. I
finished with two hours to spare, and then went back and pinged everything
from every router, fixed a bunch of mistakes. I don't use TCL scripts; my
theory is why waste your time debugging a script when you should be debugging
your lab? It takes only about 15 mintues to ping everything. Anyhow, I left
five minutes early.
The wording on some tasks was confusing. In one case, two requirements were
apparently contradictory. The proctor helped me to sort it out ultimately.
Don't be afraid to ask them questions, and don't waste an hour trying to
figure out a poorly worded question. Talk to them right away.
When I got back to the hotel, I thought I failed. 20 minutes after the test
ended, I got an email telling me to go to Cisco's web site. My friend who
passed in April waited 4 hours, so I thought I was dead. What almost killed me
was that you have to put your written test date and score in to get your
results, and I didn't know mine! I knew it was in April, so I started guessing
at dates until I hit it. I almost passed out during those five minutes, I'll
tell you. If you're traveling to the test site, BRING YOUR WRITTEN SCORE
REPORT WITH YOU!!!
OK, my advice on passing. First, know your theory. I started studying theory
in January, when I started studying for the written. I approached the written
like a research project, poring through every book I could get my hands on.
Read non-Cisco books first: Comer and/or Stevens' books on TCP/IP, Perlman's
"Interconnections", John Moy's book on OSPF, John Stewart's book on BGP,
Huitema's protocols book. I then re-enforced the concepts with extensive lab
work, dreaming up my own scenarios to test the technologies and protocols.
After I passed the written, I began a six-month lab prep program. I went
through a different technology each week, both reading theory and
experimenting in the lab. Big topics, like BGP, I spent more time on. I did
all of the examples in Doyle's books. As I worked, I took notes. In addition
to general notes, I made flash cards of obscure commands, and made a list of
"gotcha" items--things that had tripped me up or that I tended to forget to
With two months to go, and a solid background and understanding of the
technologies, I dove into sample labs. I settled on Internetwork Expert's labs
as the best, after doing a few from another major vendor. The other vendor's
labs just threw the kitchen sink into every lab to make them "tough," but they
were not carefully thought out. I strongly recommend Internetwork Expert.
Well-designed labs not only expose you to some of the "tricks" you need to
remember for the test, but they should also deepen your understanding of how
complex technologies interact. There were two or three things on the test I
hadn't seen before, and I only solved these problems because I understood the
protocols well enough to think through the difficulty. The IE labs were
crucial in this preparation. My only <minor> complaint about them is that
their solutions guide is distributed in encrypted PDF format, and I had a lot
of problems getting the Authentica software to work. I'd also recommend buying
the Cisco Press lab book (Duggan et al.) Don't even do the labs, just study
the diagrams. They're a lot closer to the real thing than IE's diagrams. I got
tripped up in the lab because I kept misreading the notation on the diagrams.
I did one or two labs a week, and did not time myself, although I did limit my
documentation to the CD. After I finished a lab I would VERY carefully grade
myself and document my errors. I re-read my ever-growing "lab errors" doc
every night. This limited the chances of my repeating a mistake, and was one
of the keys to my success. Be sure to be honest with yourself about your
capabilities. I mean, I've run into people who are on their third attempt and
still don't know the difference between MED and LOCAL_PREF. You have to be
your own harshest critic, constantly admitting your failures to yourself in
order to correct them.
A month before my test I took the Internetwork Expert Java-based mock-lab
class. I do recommend it, but only if you are well prepared. They will find
and fill the gaps in your knowledge, but if your gaps are chasms, they won't
be able to help. The four labs I did for this class were the only timed labs
I did, and I used the class to develop my time-management skills.
As you start doing labs, you will discover many technologies that you don't
know well, or maybe never even heard of. It can be intimidating, but don't
waste your time learning the nuances of Mobile IP or server load balancing
with IOS. Just focus on the core: switching, NBMA and its oddities, routing
protocols, multicast, QoS. There are too many other IOS features out there to
learn all of them. If you know the Doc CD, you can deal with those in the
lab. One thing that I did that helped with this was to make a list with the
names of weird features I didn't know how to configure, and a one or two
sentence description. I memorized this the week before the lab. That way, if
I came across one of these features, I would know its name, which is critical
for finding it on the doc cd. E.g., it's hard to find "a feature that
verifies incoming packets are coming in on the interface that they would
normally be routed out of;" it's easy to find "Unicast RPF" on the CD,
A parting thought: I see CCIE's selling their racks on eBay after they pass,
but now that I have it that's the last thing I'd do. I'm keeping my rack and
plunging into IPv6 and MPLS; after all, aren't CCIE's supposed to know
This e-mail message is intended only for the personal use of the recipient(s)
named above. If you are not an intended recipient, you may not review, copy or
distribute this message. If you have received this communication in error,
please notify the San Francisco Chronicle (chronfeedback@xxxxxxxxxxxxxxx)
immediately by e-mail and delete the original message.