CCIE lab test strategy

Unanswered Question
Sep 28th, 2007

I would like to hear the opinion of CCIE candidates on this: If the only solution I can think of is one that appears to break either the general restrictions of the lab or the specific restrictions of the task, should I go for it, or should I leave it out.

For example, if there is a general prohibition on static routes, and I think I need a static mroute to make my multicast work, should I do it?

Of course, in real life if it is the only solution I can think of, then I would do it. But then the CCIE lab is not real life.

Kevin Dorrell


I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 4 (2 ratings)
paul.matthews Fri, 09/28/2007 - 23:56

In my experience (three attempts) there *will* be a way to do it while complying with the rules. It may be a stupid way, but there will be a way.

In my labs the IP addressing was too tight to properly do what was expected, but that was for a reason! There was something they wanted to check you understood in addressing, and it could be made to work. I could imagine something like carrying multiple VLANs across a serial link to be fair game - as I said in my suggestion of how to achieve it it is a silly thing to do, but can be done.

Of course there is always the possibility that they are playing semantics, and they don't class a static mroute as a static route!

From my experience, one could ask questions to clarify, including is a static mroute classes as a static. I must add though that my lab experience is the old style 2 day lab in 1998...

Kevin Dorrell Sat, 09/29/2007 - 00:27

Thanks for the comments Paul. I have had one attempt so far, and another coming up in a couple of weeks.

I have been doing some labs with one of the training vendors, and I guess I was seeing how requirements can sometimes be ambiguous. I was wondering what I would do if that came up in the real situation. I just wanted to cover all the options.

(An NDA doesn't allow me go into detail here about the specific ambiguity that got me worried, but it did turn out that that their "correct" solution seemed to break one of their requirements.)

THanks for the advice.

Kevin Dorrell


paul.matthews Sat, 09/29/2007 - 01:37

You are welcome - best of luck on your next attempt. Your postings here certainly give the impression that you are the right kind of person!

Some of the ambiguity is down to trying not to give the answer away. It can get a little awkward at times to say what they want you to do without giving clues. For example they may ask you to configure passwords on OSPF or HSRP. If they just come out and say that, it is no big deal to configure it. To conceal what they want they need to say something like "protect your routing from rogue devices" which then give the option of static neighbours instead.

Similarly, trying to say they want GLBP without saying GLBP would probably entail describing something like " provide a resilient default gateway on the LAN that will utilise both links"

Things have changed somewhat since I last did a lab. in my time the primary employers of CCIEs were partners, and there was a requirement that a high proportion of TAC cases be escalated by CCIEs. This was when SIS98 was the main contract, and support was paid for in a lump up front on purchase of kit. That meant TAC was purely a cost centre and they wanted partners to raise as few TAC cases as possible, hence the requiremnt of CCIE. That meant as far as Cisco were concerned, the primary aim of a CCIE was to stop TAC cases. The understanding of that point was a useful bit of knowledge in one's approach to the lab. What they were looking for was someone who could stop TAC cases. Hence the presence of things like password recovery in the LAB - they got stupid numbers of cases on that, when the info was published on CCO.

Things have changed somewhat now. With the move to shared support contracts the cost is reviewed annually, and one target is the number of TAC cases raised compared to you installed base. The requirement for CCIE escalation has gone. The clever bit is that if you raise more cases, you get charged more next year. CCIE now appears to be a way of creating apparently neutral (i.e. non employees) Cisco evangelists...

By any chance have you noticed how often Cisco staff suggest raising a TAC case now?

Kevin Dorrell Wed, 10/03/2007 - 02:56

Paul, those are useful reflections, and they do tally with the feedback I am getting from the supplier side.

In my case, I very rarely raise a call to the supplier unless it is either a hardware RMA, or a real bug. I have noticed that recently they make more of an effort to find the bugs on the database before escalating them to the TAC, (which is actually a Good Thing.) They seem to have access to more of the bug database than I do, because sometimes they come up with bug references that are hidden from me.

Kevin Dorrell


paul.matthews Wed, 10/03/2007 - 04:22

Sometimes some people ar better at searching than others, or just get a lucky hit.

I know I have gone back to people almost immediately on escalating a fault to me identifying a bug, and equally I have looked at a problem, escalated to TAC and had them do the same to me, only to be able to find the bug they found on a slightly different search.

Working at a partner, I don't think I get overly special access beyond what a paying customer could get.

tperrier Wed, 10/03/2007 - 02:39

> For example, if there is a general prohibition on static routes, and I think I need a static mroute to make my multicast work, should I do it?

A static mroute isn't a static route. It's not used for routing and to direct the traffic some way. So feel free to use them (unless explicitly prohibited of course).

Kevin Dorrell Wed, 10/03/2007 - 02:46

I suppose the ironic answer to the question "Is an mroute a static route?" would be : "No, in fact quite the reverse!" ;-)

On the day I posted my original question, I had just completed a practice lab, and had wasted a lot of time because of some inconsistent requirements, so I was feeling a bit peeved about it. But that can happen in the real exam as well. I gotta get real and ask the proctor!

Kevin Dorrell



This Discussion