No Troubleshooting Style Points – But Lots of Speed Points

It’s 2PM, the network’s down, and you have the ball on fixing it. What are your priorities? Making notes on the incorrect doc and network diagrams that are slowing down the process? Noting that the interface descriptions are outdated on a few branch office routers? OK, that’s a bit facetious – the goal is of course to get the problem fixed ASAP. How you find the root cause doesn’t matter. It’s not about style – it’s about speed. For now, I’ll just go on the premise that the most common #1 goal when troubleshooting is to fix the problem as quickly as possible – which brings up a related point with yesterday’s survey. Yesterday’s survey, for those of you that didn’t see it, asked how soon you’d do a show run command on a router or switch, once you suspected that something on a router or switch was the root cause of the problem. OK, the survey is non-random, statistically insignificant, unscientific – but interesting none-the-less. 50% so far would do a show run within the first minute, and more than 75% within the first 5 minutes. So, if you buy the “speed is King” approach to troubleshooting, you might think that a lot of us think that looking at the configuration helps us debug problems more quickly. There may be other reasons for looking at the config first as well – it’s a quick way to get a lot of information in a hurry, particularly on a device you don’t look at every day. But you can certainly troubleshoot most problems while ignoring the show run command – all the same information is viewable with the appropriate show commands. But I’m curious – if you indeed do a show run pretty quickly when troubleshooting, is the primary reason that you find and fix the problem more quickly? Or is there another reason? If you didn’t do yesterday’s survey, feel free to add your opinion, and I’ll revisit it sometime next week. OK, now to today’s thought about the CCNA exams: you get speed points on the exams. I don’t mean that you get more actual points on a question you answer quickly. For the intended target audience of CCNA – people looking for jobs as an entry-level network engineer, or people that already have that job – the exam can be difficult to finish in the allotted time. So, the speedier you can answer the types of questions you’ll see on the exams, the more time you’ll have to finish, so you’ll get an opportunity to answer all questions and get all the points. There are several reasons for the time pressure, which I’ll briefly mention a few of the more significant reasons:  Simlets that don’t allow enable mode access (ie no show run command)  Testlet “questions” with 4-5 multiple choice questions – which counts as “1” in the question count  Sim questions (answered by fixing/adding configuration)  Applied knowledge multichoice – more to read and analyze before looking at the answers  Subnetting that requires finding 4-5 subnets for 1 question So, let’s say you sit down at the testing center, and you see that your test has 50 questions, 75 minutes. You get 1.5 minutes per question… but not really. If you get 2 Sim/Simlet questions, you should count on 5-10 minutes for each, depending on your hands-on skills. And a testlet question – one scenario with more than 1 multiple choice question about the scenario – counts as 1 question in that count. So, if you figure 10 minutes each for 2 Sim/Simlets, you have 55 minutes for the 48 remaining other questions. If 1 of those is actually a 5-part testlet, your per-question time on the rest of the questions is about a minute each. Then, some multichoice questions simply take more time due to the amount of reading and thinking required. For example, a question might ask why PC1 can’t ping PC2, with the problem being related to IP addressing. To find the answer, you must find the subnet number associated with 5 different host and router IP addresses, and find which one’s should match but do not. If you’ve not mastered the math and processes to find subnet numbers, you’ll certainly take more than 1 minute to answer that particular question. So, what do you see as the biggest time eaters for the exam? How much of an issue do you think it is if you get a Simlet question that requires you to find a problem, but without being able to do a show run? How much more time do you think that would take compared to if you could do a show run? Also, I’m curious if when testing layer 3, if you typically start with ping, or with traceroute? Next week, I’ll get us going on some ideas for how to get fast enough so the time pressure on the exams isn’t so bad. I’ve also been asked offline to get into the CCENT cert and ICND1 exam a little, which should be interesting. Type with you Monday if not before. Wendell

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2007 IDG Communications, Inc.