From: Howard C. Berkowitz (hcb@xxxxxxxxxxxx)
Date: Sat Jun 01 2002 - 18:50:37 GMT-3
I don't know if anyone else has had the same reaction to this thread
title, but I keep reading it as food for CCIEs. Bruce Evry has spoken
extensively of the value of blueberries, but are there other
recommendations?
More serious comments inline.
At 1:26 PM -0700 6/1/02, Jake wrote:
> I tend to agree with Matt, but I think that "networking" has
>exploded to encompass
>so many different technologies and areas of IT, that I challenge
>anyone to show me an
>"expert in networking." As far a Cisco's layer 1-4 networking, I
>still think the CCIE is
>the pinnacle.
I agree absolutely that no one can be an expert in all phases of
networking, especially
if you include network design and network product development, the
latter being outside
the scope of CCIE. I'd rather see a greater variety of in-depth CCIE
variants, because
I think that actually would better reflect industry hiring needs. At
varying levels of
one's career, I think one needs to decide on specialization. For
example, I have done
quite extensive mainframe SNA networks, but I haven't kept up with
that because my hands
are full with both current and future unicast addressing and routing.
Senior people may
also have a very valuable mix of skills that complement their
networking -- at the same
time I'm working on BGP and whatever-it-is-that-comes-after-BGP, I'm
also involved in
medical systems and stay current on medical privacy, pharmacology,
medical sensors, etc.,
and how these releate to the network.
>Obviously, it's not the end of learning, probably only the beginning...
>just like they tell you the Black Belt is only the beginning of you
>martial arts
>training. People who don't like to read or learn, have no business
>in this industry. It
>isn't like the Nobel Prize, but it really does take a lot to be a
>CCIE; since I've met
>dozens of absolutely incompetent people who hold senior networking
>positions but couldn't
>even pass the IE written (or NP many of them) I have to say that
>"experience" alone is
>not a good measure of someone's skills either. However, on the
>CCIE's relevance I'll copy
>this excerpt from another message I wrote:
> "I think 99% of the CCIE's on this group will agree that the
>CCIE IS fair. Its lack
>of relevance lies only in the fact that it tests the most difficult
>aspects of routing
>and switching and not the piece of cake "day-to-day" crap you run
>into supporting a
>production environment (which, incidentally, former secretaries are
>doing at my [Fortune
>500, 16,000 user, 180 Cat6500] company. Any 2 year old can ping, do
>a 'show port,' and
>pick up the phone and call TAC as soon as there's the slightest problem."
> Having the 4 numbers doesn't mean you know everything, but it
>does mean you have a
>very solid foundation, are willing to learn, and know where to go
>for answers if you need
>them. Plus, it says a few things about your character.
Agree completely, although I would like to see the "know where to go"
section expanded even
further -- perhaps following the written, an open book/Internet test
of whether you can find
things both in Cisco documentation and in RFCs and the like.
>
> As to Emmanuel's interviewer's quote, I know I'm going to get
>flamed for this, but
>here goes: Most people I've talked to say that the CCIE has gotten
>much more difficult
>since the early days, and because of the certification's reputation,
>Cisco has taken many
>steps and made many changes to make it more and more challenging.
>When the first CCIE got
>his number (#1024 I believe) in 1992,
I may be off, but I seem to remember it starting around 1995. I do
remember there was no CCIE program when I got my CCSI in 1993.
>he was living in a different networking world.
>Hell, they didn't even introduce Catalyst switches until a few years
>later. Back then,
>the pool of knowledgeable people was infinitely smaller than it is
>today, and the modern
>phenomena of Transcender paper-certification was probably unheard
>of. The CCIE was also
>the only certification Cisco had. In other words, the material they
>tested was material
>which the CCNP covers now and which would have been relatively less
>challanging to the
>pool of candidates today. I could be wrong, but this is how I see it
>(with an admitted
>bias, of course). ;-) Mr. CCIE #3XXX probably DID have a test which
>was relevant to
>supporting the simple RIP routers and hubs of 1994 or 5.
Definitely not. Even the first Cisco software class, RSC, had RIP,
IGRP, a little ISO-IGRP, OSPF, and assorted desktop routing
protocols. When ACRC split off, it went into OSPF, EIGRP (when
available), bridging of many flavors (translational, transparent,
SRB, etc.), more advanced desktop and IBM filtering and optimizations
(e.g., explorer caches), etc. When I taught it, I typically polled
my class about their
interests (remembering that cramming for certifications wasn't an
issue), and usually went far
deeper into BGP than the course materials. I'd often stay after class
and go into OSPF internals.
This wasn't unique, although peoples' talents vary. There are people
that are better troubleshooters than I am (Priscilla, David Desko,
Terry Slattery, Kip Peterson, Bill McInerney), where my strongest
area is in large-scale network design. Old-timers with strong
backgrounds there include Pete Welcher, Neill Craven, and lots of ISP
people not usually thought of in the general Cisco world.
> Show him a scenario with 6
>modern routers running: 5 IGP's, BGP, VoIP, ATM, QOS, DLSW+,
>Translational bridging, VLAN
>Trunking, ISDN, Multicast, (just for starters) and tell him to build
>it and make it
>converge in 8 hours... and I think he'll show more respect. However,
>I'll admit the
>market for us has definitely gone down.
There wasn't an explicit CCSI certification track in 1993 or 1994,
just what the Cisco trainers
would throw at you over a couple of weeks of observation. But I can
certainly remember putting
together 12-router networks with RIP, OSPF, IGRP, BGP, IPX
RIP/RTMP/Banyan/XNS RIP, translational bridging, IBM extensions, ISDN
or analog dialup, and BGP. For in-house training at AT&T and MCI, I
developed up to 24 router teaching labs with pretty deep OSPF and
BGP. Sometimes getting 24 (or even more) students get their routers
interoperating is harder than doing it yourself!
This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:20 GMT-3