From: Swan, Jay (jswan@sugf.com)
Date: Fri Aug 17 2007 - 19:26:25 ART
My opinion: you'll learn a lot by trying to adapt workbook labs to
whatever resources you have, whether those resources are real routers or
Dynamips. Yes, it takes up some extra time, but it's still valuable
experience.
During my final preparation phase I had access to a lab with a few
routers, two 3550s, and a 3560; and Dynamips on a Dell D620 with 2GB
RAM. I was never able to completely duplicate any of the commercial labs
with this equipment, but I was able to modify the labs to meet my study
goals and pass the exam. In some ways, I think the experience of
modifying the labs as needed was more educational than doing them as
written.
Jay (#17783)
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
scott_ccie_list@it-ag.com
Sent: Friday, August 17, 2007 3:34 PM
To: ccielab@groupstudy.com
Subject: IE Workbook Lab Topologies
Hi all,
I admit that this is a cross-post from the Professional board, but I
didn't yet have myself subscribed to this list, so I hope you will
overlook it just this once (my guess is that there are many here who
never look there, so chances are probably a bit higher someone will have
some thoughts to offer)...
____
OK all, I know that many here use these workbooks, so I'm hoping for
some insight. I had planned to use a large 8 x CPU server to run all
router instances in Dynamips. I had then thought to buy two or perhaps
four 3560-8 switches. However, when I look at the drawing on page 18 of:
http://www.internetworkexpert.com/downloads/iewb-rs.v4.00.sample.lab.pdf
(1 MB file, BTW)
I see that, for example, SW2 has connections to four different distinct
routers. I had planned to have one GBIC connection to the server per
switch (four NICs on the server). Not sure, exactly, how to work this
out without doing too much customization of each lab, which could be
confusing and chew up a lot of valuable time. The one obvious thing I
can think to do would be to create an emulated switch for each physical
switch. Then all routers would terminate to emulated switches only, per
the lab topology. I would then bridge the emulated switches to physical
server NICs using the Windows loopback and run 802.1q trunks to with the
physical switches.
For those of you who have experience with the IEWBs, do you think this
feasible? Or too much trouble to keep straight, since it wouldn't map
exactly to the lab topology.
And yes, I realize that those 8-port switches would not fully support
what I'm seeing on page 18. My thought was that rather than three links
between each switch, I'd drop it down to two in some or all places.
Thoughts on the impact of that approach?
Thanks much...
____
The bottom line is that I can't afford four 24-port 3550s or 3560s
(self-employed, slow year). So I'm just trying to work out some form of
lab prep that is not horribly expensive yet is still effective. That
server sitting there doing nothing is just too great a temptation; I
cannot justify bying a bunch of used or new routers with that resource
available to me. Just need to work out the switching part and I'm not
too keen on rack rentals. Seems like you need to book too far out and
availability can be very spotty. I'm certainly open to the idea of mock
labs, just don't want to rely on rentals for study, which I do at very
odd hours at times...
This archive was generated by hypermail 2.1.4 : Sat Sep 01 2007 - 11:32:11 ART