December 1, 2008
October 28, 2008
Mini Lab
We recently had a position available for a Senior Network Engineer and had a plan to filter the majority of the candidates out via a technical interview and shortlist two or three candidates for a thorough second interview. The idea on paper sounds pretty good so I set up a lab for the candidates to complete. This went through some beta testing at the office which was very interesting. During the beta testing in the office some said the lab was fair (those with a 80%+) while some complained that it was to easy (those who scored in the 60’s), ironically. The majority of the candidates who applied for the position already passed the CCIE written and was apparently preparing for the lab, therefore some harder questions were added to 1) make the lab more challenging and 2) make the lab interview a full hour long. Personally I think one hour is still a bit short for a practical interview, four hours would be ideal.
The mini lab consists of four devices. The candidates only have ‘enable’ access to one device where all the configuration is to be done. The other three devices have been fully preconfigured. In other words if the configuration is done correctly all links or neighbors will come up correctly as well. The candidates have exec access to the three preconfigured devices, but not access to change the configuration. This allow them to verify their configuration but also one or two questions required that the candidate do some show commands on the preconfigured devices to determine the missing part of the puzzle.
In a sense the lab interview was a failure due to the candidates failing miserably. I decided to mark the lab for each configuration individually as the questions built on each other. Full marks were awarded if the question was configured correctly even if it did not work due to an earlier mistake. Depending what the mistakes were the interviewee would score between 1 and 2 points out of 3 for each question. Examples: If the candidate knew what was needed to complete the configuration but had a wrong ip address they scored 1 point. If they configured a question correctly but made a typo or silly mistake they score 2 points. This made the scores look a bit better and did not punish ‘pressure’ mistakes to harshly. The objective of the interview was to determine the technical ability of the interviewee, therefore typo’s and real silly mistakes were a minor issue.
Interview Lab
This lab will also be good preparation for other practical interviews or even for someone in the early stages of CCIE lab preparation. What is your view? Is the lab above ccnp/ccip applying for a Senior Networking position? What would you say is a reasonable score taking into consideration the scoring method?
I will publish the Solution Guide in a follow up post.
July 29, 2008
Mobile Lab
Most have heard that Cisco has started making lab seats available through the Mobile Lab. This is a great idea and it makes sense. All you need is a proctor, workstation and internet connection and you have a mobile lab. The only down side is the secrecy it has been operating in and that it is RS only. If you dont work for a partner you probably had to fly across the globe to sit the lab. Well hopefully not anymore. The message I got from Cisco via proctors, certification support and my ‘local SE’ is that even though only partners are informed of the mobile lab, no restriction exist that only partner employees are permitted to sit the mobile lab. It is not possible to book via the conventional tool, so how can one book a seat? The only way it seems is to contact ‘your local SE or local Account manager’, thats the message from Cisco. The response I received from my ‘Local SE’ was that once the dates/seats are confirmed (which obviously means dates/seats arent available yet for my region but in the process) it should be possible for ‘candidates’ (anyone who passed the written?) to book through their local account teams. Sounds promising.
Update:
Preliminary dates for Johannesburg, South Africa (Did not manage to get dates for any other location)
November 24-28, 2008
April 20-24, 2009
July 22, 2008
Destination RS
What a roller-coaster ride the last few months have been. At the moment Im just as relieved as glad to have the digits. I’ll be lying if I say I dont want to go again, but for now I’ll be happy to enjoy some free time. Im looking forward to some “bored” moments.
Thank You’s
First of all I have to say thanks for some devine intervention. After the first attempt thinking I passed but failed I realised that I need a bit of help. I needed things to go right, I needed some favors and that is how things turned out on the second run. From day 1 everything went according to plan. Thanks Dad.
Thanks to the best supporter in the world, my best lady for all her patience and understanding. I owe her big time, more than buying something shiny can make up for.
It would also have been very difficult if it was not for my current job/managers. I think it definitely counted in my favor that my first line manager and his manager were both ccie candidates who went big in management. I think they understood what the CCIE lab requires. Every leave application was approved without any questions, a tremendous help, I genuinely appreciate this. Also for the use of company equipment, some devices that were supposed to go into production. I definitely owe my colleagues a beer for making life at work harder for them. After passing the written one manager said that the company would support me as much as possible and I can say they went beyond my expectation.
Preparation
The day I got back home from the first attempt, I looked for seats and was lucky to pickup the Dubai date. In fact I was still in brazil when I start looking for dates. I was quite fortunate to get the dubai date as I booked it within a few hours of it being dropped. I knew that if I could just maintain my level of preparedness for the second attempt and just focus a bit better I could make it. I came up with a plan to just review what I have already done, so everyday after work I reviewed vol2 labs 1-20. I would read the question, think the solution/configuration through, check the solution guide, whether I had the same idea, look whether I missed anything important then move to the next question. It took on average 3 to 4 hours per lab. If I wasnt completely sure about my solution, I would setup a simple scenario and do some testing. This worked well up until a few days before the attempt. I started to think that I need to practise some labs, as I havent done labs for almost a month. I then did the IPexpert sample mock lab, which I am very glad I did. I missed a line in an ACL, or in another question I didnt match the same routes the question asked for. I also lost two questions due to changes I made at the end or reloading the devices. Im glad I made these mistakes, because I made a decision at the end of the cisco lab to look for this type of errors. Again, Im very glad I made those mistakes in the mock lab. This was the only full lab I did before the second attempt. I also worked through the “Lab debrief” of the CCIE RS practical labs shortcut book. I would highly highly recommend working this book. Even if you dont read everything, at least work through the Lab debrief. If you look carefully, there are some differences in the way common tasks are done in this book. I would suggest using this book’s (Cisco) method. After arriving at the hotel I had two days to recover from the traveling, luckily only a 9 hour trip this time, so I started doing some IE vol3 labs. Although I didnt have that feeling you have after getting off a trampoline, which I had in Brazil, I was very tired the day of arriving and the next day. This was mainly due to the flight being during the night and not getting proper sleep on the plane. In this two days I did lab 7 to 10. It was good to get some lab practise, which I feel helped. The core lab is a good concept, but I would not recommend fussing to make your redistribution work as in the SG. One main reason is that the core labs do not have to be graded, so corners are cut. In the cisco lab, there will be very clearly stated how the redistribution should be and should not be done. The redistribution makes sense, as it also has to be graded. After the two core labs a day, I went through Michael Zuo’s notes. Good reviewing material.
Venue
Dubai is hot, very hot, as in desert hot. They say the weather is better in the winter, but I dont know, I think it will still be hot. Two things I will always remember from my trip to dubai: The crane forest and the heat. I had planned to take photo’s of the area to make the scenery a bit more familiar for the next person but got caught up in the moment and totally forgot about this. The taxi drivers all know where the “Dubai Media Center (DMC)”, “Dubai Internet City (DIC)” and “Knowledge Village” are. All three are located close to each other. Cisco Systems are located in building 10, DIC, the driver might confuse DMC with DIC, just something to note. The taxi drivers are mostly from either pakistan or india and work 12 to 14 hours a day with no days off, so to tip is not a bad idea. The 15-20 minute ride from the Ibish World Trade Center hotel to cisco cost about 45 dirhams which is about $10 usd. The hotel is located close to the highway. Apparently you get traffic on the way into the city, but Cisco is located on the way out of the city if I understood correctly. Once you get to the cisco building, if you look like you are lost the building reception would ask you “Cisco exam?” and point you to the 12th floor, after which you get into the elevator and then straight back out because it only goes to floor 4. They will confirm that it is indeed level “twelve” while giving a three finger signal, you get it. Once you walk out on the third floor there is a sign on the right that says “CCIE lab” and “use next door” which points to the kitchen/canteen area. If you are lucky someone will be there already and open for you. Else I guess the next step would be to go to the 4th floor where the Cisco reception is. Once the proctor arrives you go to the lab, no tour or the usual run down, just “Bags there, start time, end time, ok…” Thats the signal to start. He did mention that if any hardware errors are found to let him know as soon as possible. You only get the time back it takes the proctor to fix the problem and not the time it takes for you to determine that it is a hardware problem. I would suggest having a strategy for this as well. If it takes you 15 minutes to determine it is a hardware fault. You go to the proctor to let him know. You go back to your desk and continue with another task, read the lab again, whatever. 15 minutes later he comes back to let you know its fix. At the end you get 15 minutes extra. Use it to your advantage. Lunch was 5 hours into the lab. I would recommend taking some snacks with if you writing at Dubai (and Sau Paulo) as the lunch is not much. At lunch the proctor cleared some common misconceptions regarding the lab and grading. Lunch was a bit short, only 20 minutes and the lab ended 10minutes earlier. Personally I would have preferred a 30 minute break and finish on time.
Proctor comparison
Proctors do vary. Yusuf the proctor at Dubai is tight, water tight. You either get a “Yes/No” or “What you are asking me is a syntax related question, I can not answer you a syntax related question”. He reminds me of Apu Nahasapeemapetilon, for some reason. My favourite response to a question on the day was “Am I the proctor?”. It wasnt funny at the time, but when I think back about it, he has a very dry but cool sense of humor.
Lab comparison
The look and feel of the two labs were totally different. Im convinced they are designed by two different people. One also spell better than the other 
not that Im a literature giant, quite the opposite, but it was something different between the two. The diagrams also looked night/day apart. The wording was also slightly different, with a few questions adding some additional info through words e.g “users have been…, or the network manager wants to…”. This lab had one or two questions that referred to a feature within a technology that I havent configured before, but gave enough clues in the question to find it on the doc cd. Again the questions were not difficult to configure. The first set of questions were particularly ambiguous. A big difference from the first lab. The proctor could not really help here, so I went with a guess that the questions follow on each other and if they didnt give specifics, the question is probably related to the previous question. This was a guess, I could be totally wrong, but couldnt make sense of it in any other way. This was a big time waster for something I think did not test anything on the blueprint. Maybe I missed a keyword. Overall on a scale of 1 to 10 I would rate the lab a 7, two or three sections a 5/6, one section a 8, but only due to the cryptic wording.
Random thought: Why dont they put less emphasis on wording and more on configuration? I mean the best question I have seen to date was in the CCIEpractical studies shortcuts book. The question says exactly what needs to be done. I thought hard and long about it but couldnt figure it out.
This time round I spent more time reading the lab the first time. I also took care in getting my crt windows just right. Fortunately the resolution was higher than in Brazil, so I could get 4 windows aligned, the same way I did in preparation. I was prepared to change the font size from 10 to 8, but did not need to do this as the resolution was good. The workstation was sufficient. The keyboard was exactly the same rubbish they have at brazil, some logitech keyboard that does not have a dedicated “Insert” button. This was highly frustrating the first time to have to put num lock on and off when copy/pasting. SecureCRT probably has some function somewhere to change the copy/paste keys, but I havent figured that out. I practised the two days at the hotel on the laptop, in other words without a keypad, so in the lab I left the num lock off so that I could use the “0″ as the “Insert” button and used the normal numerical buttons for ip addresses/acl’s etc. This worked way better.
The Wait
After the lab if you have to take a taxi back to the hotel, the best would be to go back to the cisco reception. They have the cab company on speed dial and all the details on a pamphlet what you might need to give the operator. Most people have difficulty sleeping the night before the lab, I didnt have this problem, but for me waiting for the results was even worse. I kept thinking “what if”, what if I missed something. What if I didnt check my verification properly. What if something I did broke something else. Very tense moments. I actually labbed a scenario up and made some changes to see if what I did could possibly have broken another. Sigh of relief, it still works. Just after 3am I jumped up at the sound of what I think was a mail coming in, the results email was there. I think it helps to be half asleep or half awake when checking the mail, cause you dont really think what you are doing. I was dreading that moment scrolling down to that section where the PASS/FAIL is written, but as the window opened my eye caught two “PASSES” one for the written and another, it took about 300 milliseconds to realize I passed. What a relief. Open the score report, which only shows the number. Am I seeing right? 21500? cant be, the night before I was looking at recent numbers on groupstudy and worked out that 21500 will be issued somewhere between the day before my lab and the day after. Thats the one that was available that I wanted.
July 16, 2008
CCIE 21500
Tis done.
Just a quick post from the airport, will write a longer version soon.
June 17, 2008
Postmortem
After the lab we went to Rio for a few days to unwind, this was very welcome. If anyone travels to brazil to write the lab I would highly recommend visiting this beautiful city, even if its just for a day. I arrived back home today after another exhausting 10000km trip. I reconstructed the lab and was amazed at how much is possible to recall from the lab. I mainly did this to determine where I could have possibly lost points and to determine wether to go for a remark. I dont have my configs but I pretty much know what I did and did not do. I managed to rewrite the lab except for one question, but hopefully I’ll remember this one later. Of course If I didnt pickup a hidden keyword this is not on my reconstructed lab and Im not 100% sure about the points allocation. I managed to reverse the points from the % from the score and questions where Im sure about the points per question. I now have a good idea of how many points I needed to pass. This is in a grey area where I think a remark might do it, but due to the 0.3% fail to pass conversion rate I have my doubts. Also I would guess that Cisco would be very reluctant to change a grade. Before anyone loose their marbles, I took a read through the NDA to see if reconstructing the lab is a violation, which Im happy to say is not, as long as I dont share it to a third party, which of course I wont. That said I can say that the IE workbooks I used have all the questions I had on the lab, maybe not to the exact word but enough to lead me to believe that the IE workbooks are “based” on a mixture of various real lab questions.
About the lab:
The lab is easier than what people make out of it. I made a decision before writing the lab that I would write the exam same information whether I fail or pass. Of course I want to say that it was extremely hard withall weird and wonderful things on. To give the lab a rating of 10/10 would do my shattered ego the world of good but the truth is, it is not. Reading through the questions I have to kick myself. I failed it all on my own. In IE terms I would rate it at 6, definitely not higher than that. Yes, I missed a few key words but that was my own undoing. The questions are to the point, there is no beating around the bush: configure X to accomplish Y without using Z. Judging from the sections where I didnt loose points I can say that they dont look for configuration that was not requested in the question. E.g. The solution to one question did not make real sense, yes, I could see what they wanted me to do and thats what I did. The output from my solution would match the requirement but if a change in the topology would have had to happen, the required configuration would not be enough. I verified this with the proctor to which the answer was: “The lab is not beautiful or perfect, the questions are constructed to test a certain piece of knowledge”. Since there was no requirement to do more than just provide the output in the requirements so thats where I left it. In short, no hidden configuration needed. The hardest question in any lab is probably redistribution and this I would rate at about 7, all sections ranged between level 5 and 7. The lab is short, I checked twice to see if I have all the pages in the binder. The assessor can be considered as a 3/4 or 75% of a full lab.
About the venue:
First impression was horrible. I arrived at 7:45am, 1 hour before, as I understood, the lab was supposed to start (8:45am). Since I dont speak portuguese or spanish it was quite difficult to explain to the building reception that I have to go to the cisco offices on the 26th floor asap. Cisco employees only start to arrive after 9am, the reception arrived at 9:15. At this point I already thought that Im at the wrong place or that cisco was not aware that there was supposed to be a lab on that day. I was wondering whether they expected the lab to be on the 6th as the original arrangement and that they were not aware of the 11th date. Luckily after the reception arrived and notified the proctor things started to settle down. We had a quick “tour” in other words to show us where the restrooms are and to choose lunch. The menu is in portuguese so I would recommend the proctors choice which was quite good. After the lunch was choosen we got the usual rundown most of which we already knew. The lab started officially at 9:34. Lunch was at 12 and ended at 6:04pm. I will post a pic of the building with un update, but due to rushing to checkout and catching the next flight the photo does not say much. I stayed in the Time Othon hotel which is almost literally a stones through from the “Torre Oeste” builing which either I think means “East Tower” but might mean south tower because the sister tower is the “Torre Norte”. Other buildings around are the World trade center and the Hilton hotel. A taxi ride from the airport to cisco will be about R$107 or 43 euro and about R$120 back. It takes about 1h30minutes, but it depends on the traffic. I would not recommend renting a car in sao paulo, the people there drive like maniacs. The taxi driver couldnt speak english but understood two words “crazy driver” to which his response was “first time sao paulo”.
Some pics of the area (Note, just quick pics I took on the way out with my phone, except for the first one ofcourse which is from google earth)
 |
Cisco is on the 26th floor on the left tower. The time othon hotel is the building in the top right corner. The tower at the bottom is the Hilton hotel. The WTC hotel can also been seen at the top center. |
 |
Hilton on the left, Torre Norte in front and Torre Oeste on the right. Just a glimps of the WTC Hotel on the far right. |
 |
Another look at the Torre Oeste |
Lab strategy:
This is probably one of the biggest contributing factors to the failure. The initial plan was to arrive a day early, take as much rest as possible in the off day to compensate for any jetlag. This was not to be, someone left a parcel unattended which lead to the airport being evacuated and therefore all flights were delayed. This lead to missing the international flight to brazil and the next flight was only the next day. This also ment having to deal with lost baggage which were not heading where they were supposed to go. With all the things that went wrong upto the point of the lab I had a bit of a negative attitude, which I was aware of. I knew it was the wrong attitude but I just couldnt get myself motivated. I was physically and mentally tired. My goal hours before the lab seemed to have gone from passing the lab to just getting over with it. The night before the lab I had to remind myself that I need be alert and focus, that the lab is important, that this “getting over with it” was planning to fail and not planning to pass. During the lab I used the restroom a few times to do this as well. I took the lab way to casually, I was less stressed than I was during the assessors and some of the mock labs. The intensity was definitely lacking. I worked at a reletively slow pace, so it felt. At some point I thought I might not finish with time left to verify. That was how the lab was constructed but by lunchtime I had completed my configuration for full reachability, this was about 3:30minutes into the lab. After lunch I tested reachability and all was fine and went on the complete the lab in 5:45minutes. I had just over 2 hours for verification which I did once slowly and then came back to questions I wanted to verify with the proctor. During this two hours I didnt make any changes, big mistake, I spent time on the questions where I wanted to double check my interpretation of the question. I should have spent that time on the questions I was 100% sure I got right, because those are the areas where I lost points. Again, being to casual was a big contributor here. I did not have to use the doccd to find solutions to the answers. I used it on two questions, one to verify that this feature will work in the condition that they specified and the second to check that the option I was using would match the requirement. This was the only optional command that I havent configured before the lab.
Positives:
The lab is most definitely passable. I definitely had the right preparation. I actually think I over did it. One month before the lab I thought I have a chance of passing and I dont think this was way off. Lab is passable for anyone that really wants to pass the exam. It is not how smart you are but more a case of “have you done this before” and being as “articulate” on the day of the lab as possible. The last is probably just as important as good preparation. It is important to include the configuration that cant be verified or not critical to the solution to “make it work”.
Negatives:
Somehow I get the impression that the accuracy required in the lab is like subtracting points for spelling mistakes in a maths/science exam. What I mean is that if you get to the right answer you should receive all the points irrespective if you did the calculation in your head or used a different name to reference a variable, or in networking terms, using a difference process name.
updated
June 12, 2008
D-day
Unfortunately I didnt make it. When I walked out I thought I nailed it. The exam it self was quite easy, but I obviously missed some important detail. I cant believe it, is what I’ve been saying for the last hour. I genuinely can not believe it. I will have to accept it quickly and get a strategy to take it again as soon as possible. The test is easy, to the point, its not an IQ test, its more of a accuracy test. Thats my take on it. Why and how I lost points I have no idea. I can guess on one or two questions that I might have missed something, but on one particular section I have no idea how it could have been wrong. Looking at the score report, I have to say that I can genuinely not believe it. I will think about asking for a remark, but for now I will take a break for a few days.
Another sad part is that I travelled 10 000km to take the exam while this very week cisco has a mobile lab right on my doorstep. I found that out when the proctor asked “where you from?”. The trip to here was a bit of a desaster, due to a bomb scare we missed our flights to brazil and we arrive at the hotel at sao paulo with the lab starting in 12 hours. I had a feeling of sea sickness during the lab that I dont quite know what it was, that didnt help, but I dont think this was the reason for the failure.
On the bright side, now I have a chance at getting ccie 21337, which would more than make up for one failure. I will write some more about the lab in a few days when I get back home.
June 8, 2008
2 Days to go
Today I spent browsing/reading the doccd, nothing intense, just browsing, noting where most things are. Im not sure how well I should know the doccd, but I can pretty much find everything I configured thus far. Also came across somethings I forgot already, so I think the review was good. Dynagen has been loaded on the laptop just incase I want to see the syntax of a particular command. The cpu should handle atleast one instance. Bags are packed and ready for the long trip tomorrow.
June 7, 2008
3 Days to go
1:40am Did quite a bit of reviewing. I saw some things I forgot already, so it was a good to review. Tomorrow I will browse the doccd. I will try not to read too much as I find myself often distracted by trying to find out to much detail regarding a specific feature or technology.
Last thought regarding the Assessor lab: Im not quite sure what to think about the ability to use the assessment report only during the first 4 hours. During the first 4hours I see it as a test, why allow access to the assessment report before the time runs up. I was just thinking because I made such silly mistakes I could have easily fixed them. Just maybe it would be better to use the feature in the second session. Im probably missing some point somewhere…
Mock lab scores summary:
Mock lab 3 53 to go = 71
Mock lab 2 44 to go = 79
Mock lab 5 40 to go = 42
Assessor 1.6A 33 to go = 74
Mock lab 6 25 to go = 74
Mock lab 1 12 to go = 80
Assessor 2b 5 to go = 72
The 70’s seem to be my lucky number, but on every lab there were debatable points which I hope to squeeze out of the proctor somehow. This plus working carefully, I think its doable.
June 6, 2008
4 Days to go
6:40pm Ok did the assessor 2b today without any hassles, which was nice. After the lab I felt I nailed it, sure I passed. Then the report came and I failed, 72%. I didnt go through the answers to check what I got wrong so I immediately started to feel a bit disposed. How is it possible that after ending the lab Im sure I passed yet the score says a different story? Is it at all possible to pass? After the one hour break I went through the answer to see what I got wrong and the mood changed. All 14 points that I lost was due to reading errors. It is interesting to note that I lost all 14 points in the sections that I did 1hour after the lab started untill just after 2hours into that lab. That means in a period of about one hour to 1 hour and 15 minutes. This directly aligns with the period that I thought Im not going to be able to complete the lab. After the 2hour 30minute mark I realised I will complete the lab and slowed down. I finished with 25 minutes to spare. But this time went quick, I was able to make one change that actually added points but only covered the first hours work when the time ran out. Ok, this might sound ironic but Im glad I made those mistakes in the assessor. That is something I can learn from, I dont have to rush, I will complete the lab, do it right the first time. How I lost points:
No spoilers:
1) Didnt configure easy requirement on all devices. Unforced Error
2) Skipped an easy requirement. Completely missed the line. Unforced Error
3) Cisco hates route-maps. Ok ok, I got the point already. Avoid route-maps at all cost. Should’ve learned from assessor 1.6.
4) Questions says use a specific value and I went and used my own favorite value. Didnt not see the difference in the question e.g. If you use bbb.ddd often and you dyslexic like me, when you read ddd.bbb, they are one and the same thing. Next time I’ll be prepared.
5) Again question says use a specific value and I didnt care to take notice.
All soft issues, the route-map is again debatable. In cisco land a line calling a route-map is one “command”. In programming land that would be called, a subfunction, but I guess cisco makes the rules in cisco land.
After seeing my mistakes and knowing why I made them, I gained hope again. It is possible.
2:25am Spent quite a bit of time reviewing today. Updated the checklist, will print this out, just in case.
