Apr 202011

This weekend I took and passed Microsoft Exam 70-451 (MCITP: Database Developer 2008). I’m pleased to say my score on this exam was much higher than its prerequisite, 70-433, which I wrote about a few months ago. Much like last time, I am happy to offer my thoughts for those who might be taking this test in the future.

How I Studied

I won’t go as far as saying that I didn’t study this time around. I did do some studying, albeit not much. After reading through the list of topics covered I decided to direct my efforts towards a few areas I felt I could benefit the most from reviewing. For instance, I felt I could benefit much more from studying transaction and concurrency strategies than from studying tables and programming objects.

Honor Student Bumper StickerI did not buy any books, study guides, or practice exams. I did this because I felt I could do fine by reading BOL and blog posts about specific topics. I have nothing against those types of study materials and have benefited from them in the past, I just didn’t feel the need for them this time around. I also had the added security of a second chance as I purchased this exam as part of a Microsoft Certification Pack, which is a great way to go if you plan on taking multiple exams. Buying them in this manner gives you a discount on multiple exams as well as a free “Second Shot” for each one.

Subject Matter

For this exam I was rather pleased with the variety of subject matter covered. I thought the test matched up with the list of topics covered quite nicely and did not feel there was a disproportionate balance of subject matter like I did with 70-433. One thing I did notice was that some topics in the list did not have specific questions dedicated to them, however those that didn’t were usually referenced in other ways, such as being an answer choice for one or more questions.

It seems that the more of these tests I take, the greater my understanding of the intent of the questions. I think the test writers do an excellent job of coming up with scenarios that you’re likely to encounter in real life as opposed to the world of ideal database design. For instance, if a question were to start off with “You are designing a database to store multiple terabytes of video” my first thought would be that I probably wouldn’t design a system that way. True as that may be, it’s quite possible to start a job or consulting gig at a place that has such a system deployed and can’t easily be changed. In that case the best option may be to work with it (at least for the short term).

Test Taking Skills

Once again I can’t over-stress the importance of test taking skills. Unless you are a complete genius, there’s probably going to be at least one question on this (or any) test that you don’t know the answer to. At this point your options are:

A) Leave the question unanswered
B) Pick an answer at random
C) Use your testing skills to eliminate incorrect choices and select an answer from those remaining

Needless to say your best chance of answering correctly lies with option C. When I’ve needed to do this, I’ve always been able to eliminate at least two choices. Here’s an example:

Q: Rachel’s query is the bestest EVAR. What hint did she use when accessing the docHistory table?


We know nothing about Rachel or her query other than it’s on docHistory, so it’s a guessing game at this point. One thing we do know is that NOLOCK and READUNCOMMITTED are equivalent, so A and D can’t be the answer. On top of that RECOMPILE is not a table hint but a query hint, so C is out as well, meaning the answer must be B. Even if you had no clue the READPAST table hint existed, you can arrive at that answer by elimination.

What’s Next

After passing this test I have now completed the prerequisites for the Microsoft Certified Master: SQL Server 2008 knowledge exam. I’ve already had a few people ask me if I’m going to take it. HECK YES! It’s definitely on my list of things to do, but I will be taking some time to prepare for it first. I have no doubt in my mind that I will get there and will work hard in doing so, the only question now is how long it will take me.

Jan 042011

What do you call someone who graduated last in their class? In this case, me! I recently took Microsoft Exam 70-433 (MCTS: Database Developer 2008), and am pleased to report that I passed. I’m happy to offer up some of my thoughts as well as tips for those who might take it in the future.

Scoring – Why Bother?

So you know how I just said that I passed? That’s a half-truth. The other half is that I exactly passed. The minimum passing score was 700, and that’s what I got. I feel pretty dumb, however I’m sure I would feel a lot worse were my score 699.

I’ve never been a fan of Microsoft’s scoring method for these exams. They never say what the maximum possible score is, nor is there any quantitative breakdown of how you received the score you earned. 700 is passing, but out of what? 900? 1000? They include bar graphs on the result sheet for each of the major subject areas, but I’d still like to see some numbers. For the results Microsoft provides and the almost nothing that can be gleaned from them, they might as well just skip the numeric score altogether and simply return whether you passed or failed.

How I Studied

Dunce HatMy study method was incredibly simple – I didn’t study. Now I know if my mother reads this she’s going to tell me how it serves me right for just barely passing if I didn’t study, but hear me out for a second before sippin’ the Hatorade, ok? There’s been lots of good debate over the usefulness of certifications compared to real-world experience. Brent Ozar (blog | twitter) has a great blog post about how certifications are icing, but not the cake. Anyone can buy the study kit, lock themself in a room for a while and pass the test. I was more interested in seeing if I could pass the test based on what I know right now without doing any cramming. After all, if I really know my stuff, studying won’t be necessary, right? So I purposely did nothing other than read over the list of skills measured on the 70-433 exam page at Microsoft Learning. While I did pass, I probably won’t use this exact same method again in the future and will instead try to find a happy medium between cramming and no preparation at all.

Subject Matter

Another thought about that “skills measured” page I linked to above: I’m not sure how accurate it is with respect to topic weightings. The test I took had 55 questions. There’s an awful lot of topics and skills listed on that page, and I realize it’s rather unlikely that all of them will be covered in only 55 questions. From my perspective, it sure seemed like there were a disproportionate number of XML questions – far more than the 12% listed there. 12% of 55 is 6.6, and while I didn’t keep count, I’d be willing to guess that probably 10 questions dealt with using XML-related functionality – most of them in ways that no sane DBA would bother. In the same way that owning a hammer doesn’t mean that every problem is a nail, SQL Server having XML features doesn’t mean that everything to do with XML should happen in the database. Application servers make XML parsing/searching/manipulation a lot easier in many cases!

Eliminating Stupid Answers

To me, test taking skills are things like keeping track of time, pacing oneself, knowing when it’s appropriate to skip a question and move on, and eliminating stupid answer choices – pretty much everything you can (legally) do during the test to help improve your score without actually knowing the material. Good test taking skills can really save your tushy at times, and Microsoft exams are no exception. I’ve found there to be a fair amount of just plain stupid answer choices in Microsoft exams, to the point where I’ve made a game out of finding and eliminating them. I usually don’t have much trouble dropping 2 choices from each question. If you can do this, your chances of choosing the correct answer greatly improve. Would you rather pick a random answer from 4 choices or from only 2? Here’s a made-up example and how I’d go about solving it:

Steve is querying the “Widgets” table for the top five widgets in terms of sales. What syntax should he use?

A) SELECT TOP (5) FROM Widgets ORDER BY Sales;

The first thing I always do is play “which of these is different?”. In this case, B screams out to me because it’s selecting the top 10 instead of the top 5. The question’s asking for the top 5, so B is out. Similarly D is also different, as it’s selecting from the Tools table – everything else is selecting from Widgets. Once again, referring back to the question, we know that D can be eliminated. Now the only remaining choices are A and C. Even if you’re clueless as to the meaning of the “DESC” keyword, your chances of choosing correctly are twice what they were before.

All in all, this exam was alright. I wish I would have done better, but am happy that I passed. Next up for me will be exam 70-451 (MCITP: Database Developer 2008) which I hope to take in April or May. I’ll be sure to write about it when I do.