Re: Success

Hi again all,

On Thu, May 28, 2009 at 10:51 PM, Dale Shaw <dale.shaw_at_gmail.com> wrote:
> I passed.
>
> CCIE #24464 -- R&S.
>
> I am very happy and quite relieved.
>
> More later, I have a hangover to work on.

As promised, here's a bit more background.. I hope you can focus for
long enough to read the whole way through. This e-mail is almost as
long as my CCIE lab notes.

I've been in the game for 13 or so years -- started at the beginning,
building and repairing PCs. I still think the troubleshooting
methodology I developed in this phase has been invaluable. I've worked
for systems integrators in the channel the whole time, mostly in the
systems (Microsoft / Novell / Citrix and UNIX) infrastructure arenas.
I'd always had a keen interest in comms from my early years as a
teenager BBS sysop (RemoteAccess/Pro on everything from NetWare to
Desqview to OS/2). Around 10 years ago I inherited the responsibility
for a cisco 4500M and the company's Internet connection. I was pretty
chuffed when I got multilink PPP working over basic rate ISDN, despite
taking us off the air and locking myself out of the device several
times in the process. I learned about the implicit deny at the end of
ACLs this way :)

I passed my CCNA exam in the year 2000. At the time I was responsible
for a large Avaya (Octel/Lucent) TDM-based unified messenger
installation based on Unified Messenger (very cool for its time). I
had more experience and interest in the messaging (Exchange) side than
on the voice side. I sometimes wonder how things would've turned out
if I'd dug into the voice stuff a bit more.

In the years that followed I'd been given several opportunities to
install and configure cisco PIX firewalls (the 'Classic' and 500
series devices) and dabbled with IPSec VPNs. By this stage I was still
working for a systems infrastructure outfit but we had the
VPN/Security partner specialisation so we got the occasional gig in
that market. Around this time I passed the exams to become a Cisco
Qualified Specialist (CQS) in Security. This cert later morphed into
CCSP and I think I just had to pass one or two more exams for that.
That was 2004 or so. I also started working on CCNP, which was a great
learning experience. I used the Cisco Press self-study set. Going
through this process filled in lots of gaps and exposed me to some of
the technologies used in larger enterprises. I failed BSCI first go
but got it second time around. Up till then, despite being a MCSE
(since '98, when it meant something), CNE (since '99) and a handful of
other less meaningful certs, I had never failed an exam. It was a rude
shock! :-) CCNP was wrapped up in the same year, 2004.

In 2006, despite enjoying the work environment with the company I was
employed by, I made the decision to focus on networking and I needed
to move. I'd just finished a large Exchange implementation, which was
fun, but my heart was at the command line :-) I switched employers and
started working for one of Australia's larger Gold partners. It was a
network design gig, but in reality there was/is a great mix of
hands-on and design work. It's been a long time since I installed a
Microsoft service pack for a customer, and while I sometimes miss
being a 'systems' guy, I consider that a good thing!

I originally booked my CCIE lab in Sydney a year ago (May '08), having
passed the written in March, but I moved to Canada for a working
holiday and with all those novel animals like bears, moose, squirrels,
skunks etc. there was never going to be enough time to prepare -- I
was kidding myself. I first postponed to September '08, then January
'09. I returned to Australia in mid-November and realised I had a
unique opportunity to give this thing a real go. I booked myself in to
Narbik's January boot camp in Sydney and started working my way
through his "Soup To Nuts" workbook about a month earlier. Attending
the boot camp was a real eye-opener for me. Up till that point, I knew
how much I knew, but not how much I didn't know. I didn't get as much
out of the boot camp as I could have if I was better prepared, but it
was critical to setting my focus on the end goal. I met some nice
blokes that week -- some of which are occasional GroupStudy
contributors (hi guys!).

For the next couple of months, I worked through Narbik's advanced and
boot camp workbooks. I wanted to complete them from start to finish
before locking in a lab date. When I finished pass #1, I booked and
paid for my lab -- April 21st. I lined up some ASET lab sessions
through my local Cisco SE and was pretty comfortable with the content
and happy with the results I was seeing, noting that they're scored
differently than the real lab and the wording of the tasks tend to
steer you toward the 'right' solution. I was only working part time
throughout my study, so I was able to dedicate about 20-30 hours per
week to study without too much trouble. I don't know how anyone does
it with a family and/or full time work.

Leading up to attempt #1, I was already starting to suffer from 'study
fatigue'. It's hard to stay motivated for so long, as I'm sure many of
you know. Nevertheless, I was quietly confident, and went into the
first attempt thinking that I had a real chance. Apart from the
occasional case of 'butterflies', I wasn't all that nervous. I stayed
at the Urban Hotel in St Leonards and got allocated room '224', which
I thought was pretty cool (it's on the corner facing the Pacific Hwy,
so take ear plugs if you're a light sleeper!). Even walking down to
the office, I was feeling good. I had that "well, there's nothing you
can do to improve your chances now!" feeling, which was comforting. I
guess that feeling could work for or against you.

There were three candidates that day, including me -- one for SP, one
for Voice and me, for R&S. We got our name labels and made our way
into the briefing/lunch room. There's a CCIE surfboard in there, for
some odd reason. I guess the marketing department had some extra dough
to throw around :-) I was still feeling good -- no nerves. Once the
briefing was over and we moved into the lab room, I can't explain why,
but my heart rate started to increase and the nerves kicked in good
and proper.

First I had to log in and take the OEQs. Man, I was so relieved when I
saw them -- they were easy! I didn't time myself, but I guess I took
10-15 minutes working on them. I was as verbose as I felt I needed to
be to indicate my understanding of the content. Some questions needed
more; others needed just a short sentence. When I clicked 'Submit', I
had no doubt that OEQs would not be a problem for me that day. All
I'll say about the OEQs is this: IF you have a chance at passing the
configuration portion of the lab -- i.e. you've done the hard yards,
learnt the technology and practiced enough -- the OEQs will not
challenge you. It's as simple as that. I know some people pass the
configuration section but fail based on OEQs and that flies in the
face of what I've just said, but that's my view. My hat goes off to
those of you who speak English as a second (or third.. or fourth)
language. Amazing.

Once you've done the OEQs, you need to log out and back in again using
a different set of credentials. This is all made clear to you, so you
don't need to worry about the details -- pay attention though;
apparently it's not uncommon for candidates to approach the proctor
with login problems. Once you've logged back in, you're able to access
your lab equipment and begin the configuration portion. Reading
through the workbook, I didn't see any "gotchya!s", but by this time,
I was really on edge. I shouldn't have had that coffee at the hotel
(although it was surprisingly decent) because the adrenalin was
pumping so hard I needed no supplementary stimulation. I started
working through the material..

Cut to just before lunch time, and I had my switching and IGP sections
under control. I had full IGP reachability and I remember how I'd read
this would be a reassuring feeling -- I guess it was. I'd also read
that it was good to be starting on BGP around this time, but
everyone's different and take it from me, not every lab has BGP coming
straight after IGP anyway. For some reason I'd ordered two sandwiches
in the morning because I envisaged those cocktail sandwiches turning
up on a catering plate. That's not the case. Somehow the lunch orders
turn up and they're full-sized sandwiches. I ate the tuna (brain food)
sandwich and left the chicken schnitzel wrap till later. It would be a
~AUD$2,200 wrap devoured later that evening.

After lunch, and even before, time was moving fast. I kept glancing up
at the clock and I thought I had time under control, but I did kill
some time renumbering an OSPF area ID after realising I had transposed
two digits in the area ID (!). I literally had to change the config on
all affected devices in area ABDC to area ABCD and re-verify
adjacencies and reachability. By the time the end of the day rolled
around, I thought I had met all requirements other than one IOS/IP
Services task that I didn't commit any config for 'cause I couldn't
figure it out -- one of those ones you sacrifice because you know
getting it wrong could break other things. One problem at this point
-- I had _almost_ full reachability to backbone networks, but not
quite. There was a subset of the backbone-learned routes that I could
not reach. I decided to leave it be, as there was no explicit
requirement to be able to ping a host address in each backbone prefix,
but this would prove a mistake in strategy/interpretation.

I walked out of the building thinking it could go either way. I knew
I'd given it a good nudge. I also knew the result would come
reasonably quickly, because for R&S at least, the lab is graded by (if
I recall correctly) the Belgium/Brussels office. I walked back up the
road to the Urban, got changed, checked my e-mail once or twice, then
set off to find some beers in Crows Nest (about 2km down the road).
I'd pre-configured Gmail to apply a 'star' to anything from
"ccie_at_cisco.com" so it would stand out when it arrived. Back to the
hotel.. Like a lot of people, when I opened that e-mail (at about
7:40pm, about 2.5 hours after leaving the lab), my eyes were drawn to
the "PASS" text in the e-mail itself. Do not make the same mistake --
your result is NOT included in the e-mail itself; you must log in to
the CCIE site to see the result. Anyway, my heart sank when I saw
"FAIL" on the web site, and I reviewed the score report. It had
highlighted some areas that I suppose I knew deep down were weaker --
multicast, IOS/IP services, and BGP. I knew, with more time, I would
have ironed out the kinks in my topology. I guess everyone says that
:-) I'd rate the difficult level about 7-8 out of 10, and the biggest
surprise (to me) was how heavily configured the lab equipment was at
the start of the day. Don't ask me any more about this, just know it
was a surprise, and something I'd never read about prior.

I was disappointed, sure, but I had low expectations for attempt #1. A
day or two later I was back home, and I had booked my 2nd attempt for
May 28th. No mucking around. I decided I'd work on my weak areas
(noted above) and see what happened. I knew I didn't need to radically
change my prep, as I'd come very close to passing 1st attempt. Time
management was a bit of a concern, but I thought I was already
reasonably snappy, so I didn't focus on this aspect of the failure. I
read 'Internet Routing Architectures' for the first time, devoured as
much IOS/IP services (mainly NAT) content I could find, and worked on
AutoRP and BSR configurations -- my foundation knowledge of multicast
was/is sound.

The month (or so) between 1st and 2nd attempts was really hard.
Despite having the 2nd attempt booked and paid for (and BTW this is
all coming out of my own pocket), I was really struggling to get
motivated. It didn't help that my day job was generating a significant
load for 2 or 3 weeks leading up to the lab date. I probably only
managed about 40 hours' study for the whole month. I don't know if
this helped clear my mind or what, but obviously the result was better
for me this time.

Attempt #2 was a completely different day. Again there were three of
us -- 1xSP, 1xVoice, 1xR&S (me). I felt that the lab content was
slightly easier -- the conspiracy theorist in me wants to believe that
Cisco does this on purpose but who knows -- but maybe it seemed that
way because I was so much more composed. I didn't have a coffee that
morning, but the hands were still cold and the mouth was still dry and
the need to take a leak every 30 minutes was still there; just not as
strongly. The OEQs, once again, were not particularly challenging. One
of the four questions asked what a particular interface command did,
and to be honest I'd never used that particular parameter on that
command. I made an educated guess and later verified my answer to be
technically correct but possibly not precise enough. I'll never know
if I got 3/4 or 4/4 correct, but I had no doubt about the other three.

I had, I'd say, about 2 to 2.5 hours left over for verification. I
didn't _really_ need it, 'cause everything seemed to work well first
time. I'd rate the difficulty level about 6 to 7 out of 10. The level
of pre-configuration was minimal compared to attempt #1, but I was
ready for that anyway. At about 3pm or so I was convinced I had this
one in the bag, I even caught myself smiling thinking how good it was
going to be to pass this milestone. I re-read through the workbook and
feverishly re-executed my TCL scripts. Smooth sailing!

BUT -- and I'm not exaggerating -- 10 minutes before time was up, my
backbone connectivity disappeared completely. Routing
peering/adjacencies and the associated routes disappeared, and I
couldn't ping the backbone routers. I had 'incomplete' ARP cache
entries. Gone! I panicked, thinking that despite reloading a couple of
times throughout the day, I had a L2/bridging time bomb in my config.
I could still see the backbone devices as CDP neighbours, but I
started to un-do some of the configs most likely to cause the problem.
Thankfully I remained composed enough to remember hearing about how
other people had made drastic changes towards the end, so I
re-committed the 'working' configs. Still, the backbone was gone. The
proctor had called "write configs!" a couple of times already, so it
was getting pretty obvious that I was doing more than writing configs.
Time was up. In a flap, I issued one final reload of all of my
devices, in the desperate hope that if it was a time bomb in my
config, it'd lay dormant long enough to make it through the grading
process -- after all, it had all been working for several hours by
now.

I mentioned the problem to the proctor and he had a quick look over my
shoulder. Once we'd handed back our workbooks and left the lab room, I
mentioned again how concerned I was about what I'd seen in the last
10-15 minutes. The proctor casually mentioned how sometimes -- once
every 2-3 months or so -- candidates in other locations can screw
things up badly that it affects the shared lab infrastructure,
including the backbone gear. He mentioned how *he* would've verified
it if he were in my shoes, but of course it was too late for that now.
He said he would look into it. I asked if it would be possible for him
to immediately confirm/deny whether the backbone was affected by
another candidate, so I could leave the building having some idea of
my chances. He declined. I literally went from ultra-confident to
seriously unsure in the space of 15 minutes.

I left the building and went in search of amber-coloured solace. The
more I thought about what happened, the more I thought it must be as
the proctor described -- a meltdown caused by another candidate.
Applying logic to the situation and going through my configs in my
mind, a time bomb just didn't make sense. Network devices don't behave
in unpredictable ways like that, and the only 'fancy' bridging config
I had applied was to a completely separate part of my topology. Still,
I couldn't hold back the thought of a 3rd attempt from entering my
mind.

Fast forward a few hours -- a significantly longer wait than attempt
#1 -- the 'starred' ccie_at_cisco.com e-mail arrived. By this stage, my
girlfriend had arrived at the hotel (she'd driven up to meet me). I
said "it's arrived", and anxiously clicked the link to check my
result. I logged in and my eyes scanned to the table that had
previously contained a blank field. My girlfriend was already
screaming "PASS! PASS!" but she was reading the result for the written
exam. Once I saw "PASS" in the required location, I stood up, hugged
and kissed her, and then jumped around the room like a juvenile
delinquent yelling "two four four six four! two four four six four!".
Woohoo! It was indeed a pass :-) The overwhelming feeling was not one
of accomplishment, but of relief. The prospect of a 3rd attempt was
gone. By the way, I did work on a hangover that night, and I did a
great job :-)

Materials used:
- Real life experience in multiple technical disciplines (you can't
take a boot camp for this).
- Narbik's "Soup To Nuts" workbook
- Narbik's "Advanced CCIE" R&S workbooks (from the boot camp)
- Cisco ASET labs -- talk to your local SE
- Routing TCP/IP volumes 1 and 2 (Doyle) -- multiple times; seriously,
there is implementation-specific detail in here you won't find
anywhere else
- Cisco QOS Exam Certification Guide (Odom) -- the QoS bible
- Internet Routing Architectures (Halabi)
- CCIE R&S Exam Certification Guide (Odom)
- Cisco BCMSN Exam Certification Guide (Hucaby) -- for switching
- DocCD / Cisco.com to fill in the gaps/to beef up on certain technologies
- TCP/IP Illustrated -- dated, but great for conceptual understanding
of what's happening at layer 4 and above
- Partner E-Learning Connection (PEC) -- this is a gold mine! If you
work for a partner, don't overlook this resource
- GroupStudy subscription w/Gmail -- the ultimate searchable CCIE R&S
lab prep archive!
- 1 x freeze-fried beard of Zeus, available in Cisco MarketPlace.

My advice:
- There are no shortcuts
- Don't rush into the lab booking. With the 90 day system now, you can
book it when you're ready.
- Make use of the partner resources if you're eligible
- Don't sweat the OEQs (I know it's hard not to and you won't believe
me till you've seen them with your own eyes)
- Done properly, pass or fail the CCIE journey will make you a better
operator. It will hone your precision and enhance your ability to
think clearly under pressure.
- Search the GroupStudy archives, trim your replies, avoid
top-posting, don't inhale, book a hotel near a pub, don't stay in
Sydney any longer than you need to, because it smells funny.

Shout-outs:
- I'd like to thank my girlfriend, who was amazingly supportive
throughout the whole process, and my imaginary friend, Ferdinand. Just
kidding, it was mainly Ferdinand. Just kidding.
- I'd also like to thank all the regular and transient contributors to
GroupStudy, and Charles Darwin.

cheers,
Dale

*click*

Blogs and organic groups at http://www.ccie.net
Received on Wed Jun 03 2009 - 21:40:26 ART