April 2, 2010 at 4:40 pm
Steve Jones - Editor (4/2/2010)
WayneS (4/2/2010)
where the MCITP is geared towards a specific version, I agree that this certification should be Version specific.Definitely version specific, though I'd also add I think it needs to be area/subsystem specific as well. No one is a complete journeyman in all aspects of SQL Server.
Agreed.
Like with the trades, a typical journeyman is not a journeyman in plumbing, electrical, cabinetry etc. They usually pick one. That's not to say that a person couldn't certify as a journeyman in each of the different emphasis areas - just that I wouldn't expect a journeyman electrician to know plumbing unless he was also a journeyman plumber.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 2, 2010 at 5:13 pm
WayneS (4/2/2010)
Grant Fritchey (4/2/2010)
I'd say the cert can be tied to a product version. Say you get certified for SQL Server 2000. You'd still be fully employed here where I work and your cert would be good. Another place may only have need of people who know 2008 R2, so that 2000 cert would be worthless. But I don't think you should have to reaquire the cert just because time has passed.In view of the way that MS currently does it (
Credential - Certification/Version
MCDBA - SQL Server 2000
MCTS - SQL Server 2005
MCITP - Database Administrator)
where the MCITP is geared towards a specific version, I agree that this certification should be Version specific.
I don't have a problem with the TIS proposed by Wayne, but I'm going to be the voice of dissent on being version specific. Sure at the time of original certification the test will cover most current version, but maintenance should be through continuing education not testing. So the cert would be MCJ - SQL Server DBA and then you maintain that certification through continuing education credits. Similar to teachers and doctors. Things change in those fields and you keep up with continuing education.
Jack Corbett
Consultant - Straight Path Solutions
Check out these links on how to get faster and more accurate answers:
Forum Etiquette: How to post data/code on a forum to get the best help
Need an Answer? Actually, No ... You Need a Question
April 2, 2010 at 5:23 pm
WayneS (4/2/2010)
Man has this topic been hot today! I find that fantastic! (I almost wish I hadn't gotten off work early.)I feel that the Time-In-Service (T-I-S) has been fleshed out pretty well (from what I see, the only real discussion is on the amount of time), and I'd like to propose this as the T-I-S requirement:
The candidate attests to a minimum of 18 months hands-on experience in SQL Server. Willfully misrepresenting (lying) about this is subject to a permanent revocation of this certification.This does NOT mean an equivalent of 18 months of 8 hr days with hands-on experience; this means 18 months of any hands-on experience. This can be as a developer, DBA, or in BI (SSAS/SSIS/SSRS).
The 18 months is based on a compromise between the 1yr and 2yr being mentioned as the most frequent requirement. This also greatly minimizes any administrative burden from any certifying agency in the verification of this.
I wish that there was a way, from within an existing thread, to do a poll. Barring one, can we have a simple Yes/No vote by replying to this message?
YES!! I find this totally reasonable. I dont have any issues with the certification being version specific although Jack's proposal of added credits might be make it more attractive, since people take a *long time* to upgrade and that might make it a sort of a race with little practical experience.
April 2, 2010 at 5:26 pm
I just found this comment from Joe Sack concerning the MCM.
The next large update of MCM will be for the version after SQL Server 2008 R2. Those with SQL 2008 MCM certification will be able to upgrade their existing cert via an exam/lab combo.
Thus for the MCM (MCM SQL 2005 is retired), you can update via a lesser protocol than the initial certification (kinda like MCDBA from 2000 to 2005 where a minimal set of tests was required).
I could almost see that a single test to upgrade or maintain my certification could be like continuing education credits. If there was an option to do either the test or provide evidence of continued education - I think that would be the optimal method for this requirement.
I think evidence of continued education would involve Presenting, panel review, or having written articles/books that have been published. There should also be a nominal fee involved.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 2, 2010 at 8:21 pm
Jack Corbett (4/2/2010)
I'm going to be the voice of dissent on being version specific. Sure at the time of original certification the test will cover most current version, but maintenance should be through continuing education not testing. So the cert would be MCJ - SQL Server DBA and then you maintain that certification through continuing education credits. Similar to teachers and doctors. Things change in those fields and you keep up with continuing education.
CirquedeSQLeil (4/2/2010)
I just found this comment from Joe Sack concerning the MCM.The next large update of MCM will be for the version after SQL Server 2008 R2. Those with SQL 2008 MCM certification will be able to upgrade their existing cert via an exam/lab combo.
Thus for the MCM (MCM SQL 2005 is retired), you can update via a lesser protocol than the initial certification (kinda like MCDBA from 2000 to 2005 where a minimal set of tests was required).
I could almost see that a single test to upgrade or maintain my certification could be like continuing education credits. If there was an option to do either the test or provide evidence of continued education - I think that would be the optimal method for this requirement.
I think evidence of continued education would involve Presenting, panel review, or having written articles/books that have been published. There should also be a nominal fee involved.
What I'm mentally putting together is:
MCITP is version-specific (can upgrade to next version by an upgrade exam).
MCM is NOT version-specific, maintained by upgrade (exam/lab)
Since MCJ-DBA/BI is to be a stepping-stone between the two, it should be maintained by continuing education or an upgrade test (dissent about whether to have an upgrade test...). (This is a change on my opinion from earlier, based on how the MCM is maintained.)
I like the idea of presenting or writing (obviously about the new features). I also like the test. But the two do not seem, well, "equal" to me... a test would cover many (if not all) of the new features, and the presenting or writing would likely only cover just one or a couple. This just doesn't seem fair to me. To be fair, the continuing education piece would have to cover most of the new features.
It may be that we need to shelve re-certification for a while, until we hash out what is needed to get the certification in the first place. Then we can decide what needs to be done for re-certification.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
April 2, 2010 at 9:08 pm
If it's not version specific, then you'd need to note which version you passed on , and also have documentation of CE for future versions.
April 2, 2010 at 9:10 pm
Heh... I just read Jason's blog about this thread and he talked a little about the price of getting an MCM.
Certification just shouldn't cost that much... it excludes people that may have the intelligence and just not the bucks to invest. It's also the reason why a lot of smart folks get left behind in higher education... too much money and not enough educating. Heh... does anyone know if MCM's are taught how to use something like a Tally table or how to do a dynamic cross tab in their 3 weeks of bootcamp?
--Jeff Moden
Change is inevitable... Change for the better is not.
April 2, 2010 at 9:12 pm
I haven't gone through Brent's Blog on it, but I doubt they have a lot of T-SQL taught. I think it's more details on the SQL Server core product, behaviors, tips, tricks, switches, etc. on the internals and not necessarily writing great code.
I do think that T-SQL ought to be a separate cert from core SQL Server
April 2, 2010 at 9:28 pm
Gosh... :blink: You would think that an MCM would be required to know how to distinguish between good code and bad code as part of some sort of "performance tuning" requirement.
I guess that brings us to the subject of this thread... what would be the "core" curriculum of the proposed MCJ cert or could it have multiple paths? Perhaps a "Systems DBA" path and another for "Application DBA"? Maybe even a 3rd path for "Journeyman Developer" or "Hybrid DBA" (one that was close to "Master Developer" with a whole lot of systems knowledge but not a full blown "Systems DBA")?
--Jeff Moden
Change is inevitable... Change for the better is not.
April 3, 2010 at 12:49 am
Jeff Moden (4/2/2010)
Gosh... :blink: You would think that an MCM would be required to know how to distinguish between good code and bad code as part of some sort of "performance tuning" requirement.I guess that brings us to the subject of this thread... what would be the "core" curriculum of the proposed MCJ cert or could it have multiple paths? Perhaps a "Systems DBA" path and another for "Application DBA"? Maybe even a 3rd path for "Journeyman Developer" or "Hybrid DBA" (one that was close to "Master Developer" with a whole lot of systems knowledge but not a full blown "Systems DBA")?
That surprises me too. Seems there is a gap in the market for a 'I write good code' certification.
In my mind, there has always been a distinction between Administering a database, and Developing one. From my viewpoint, administration is about configuring SQL Server, keeping it running, ensuring there is recovery strategy, monitoring performance, and so on. Developing/architecting is more centred on concept, physical/logical design, and code-writing.
Database people that can perform both (partially overlapping) roles competently seem comparatively rare...?
April 3, 2010 at 2:04 am
Steve Jones - Editor (4/2/2010)
WayneS (4/2/2010)
where the MCITP is geared towards a specific version, I agree that this certification should be Version specific.Definitely version specific, though I'd also add I think it needs to be area/subsystem specific as well. No one is a complete journeyman in all aspects of SQL Server.
Absolutely. I get almost daily, personal, reminders of this one.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
April 3, 2010 at 2:07 am
Steve Jones - Editor (4/2/2010)
I haven't gone through Brent's Blog on it, but I doubt they have a lot of T-SQL taught. I think it's more details on the SQL Server core product, behaviors, tips, tricks, switches, etc. on the internals and not necessarily writing great code.I do think that T-SQL ought to be a separate cert from core SQL Server
I think it would be interesting to see how much scripting is done. IIRC, Brent said they had 100's of scripts to review for the exam. Not sure what the details of those scripts were though.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 3, 2010 at 2:10 am
Jeff Moden (4/2/2010)
Gosh... :blink: You would think that an MCM would be required to know how to distinguish between good code and bad code as part of some sort of "performance tuning" requirement.I guess that brings us to the subject of this thread... what would be the "core" curriculum of the proposed MCJ cert or could it have multiple paths? Perhaps a "Systems DBA" path and another for "Application DBA"? Maybe even a 3rd path for "Journeyman Developer" or "Hybrid DBA" (one that was close to "Master Developer" with a whole lot of systems knowledge but not a full blown "Systems DBA")?
Going off of Brent's blog for this, but he said that it seemed the core of the MCM was centered around knowing why one solution is better than another and to know the reasons around "it depends."
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 3, 2010 at 8:05 am
If you read the goals of the program, that's what it is. Make effective decisions around design and deployment and operations. Really an "it depends" type of cert.
Which is good, we need that, but I'm not sure we need it as a set 3wk, $25 cert. To me this is like a 2 yr grad program, which might cost $25k, but something more like an executive MBA that teaches you across time.
I'm not a big believer that you can cram that much information into a short period for most people. Some of those guys might be able to, but my guess is that they'll forget a lot of what they learned if they don't use it quickly.
April 4, 2010 at 5:34 am
GSquared (3/23/2010)
I would very much favor an apprenticeship system. They have a very strong proven track record of working well (they have been the main training technique for most of the history of civilization). Certification based on testing actual ability to actually do things would make sense to me. Certification based on exams, even tough ones, doesn't add up to much for me.
I agree that an apprenticeship system is better than anything industry certifications (or many universities) can provide today. But I think exams are needed too. For certification as a journey man, I would expect the candidate to sit a written exam consisting of 5 sessions of 4 hours each, with no multiple-choice questions but questions that require the candidate to write answers that demonstrate an understanding of the trade and a thorough knowledge of the basics (answers consisting of long waffle quoting all the right buzzwords and using the right jargon don't get any marks, or maybe get negative marks - waffle and bluster are damaging not useful) plus a practical exam of about the same size; and only if both thiose exams are passed would it be possible for his master to certify that he had the knowledge, skills, abilities and attitudes required of a journeyman in the trade. For certification as a master I would expect written and practical exams of the same length but requiring a higher level of skill and knowledge, plus a requirement to provide a dissertation (sufficiently good to be considered the work of a master, but it doesn't mater whether it's a survey or research or a mixture) on some aspect of the trade and undergo an oral examination 4 hours long by his master plus three other trade masters on the subject of the dissertation (the three other masters would also see the the written exam answers and the practical examination output) and all four examining masters would have to be convinced that the candidate had reached master level. Using examinations like this turns them into a very useful tool in assessing whether someone is up to the standard required.
When my dad was in high school, they had a class that had the specific purpose of preparing the students for their standard tests (I think it was the SATs, but I wouldn't swear to that). They all had scores that were well above average, because of the class. Did that mean they were actually better educated than their peers who didn't take that class? Not if by "better educated" you mean "better prepared for adult life as productive members of society".
This just demonstrates that the exam in question (SATS or whatever) was an extremely badly designed exam. Probably multiple-choice, or maybe extra marks for waffle.
If, on the other hand, someone recognized as a master of the craft were to sign a paper saying that the person had adequate skill and experience to take on the role (as per classical apprentice-journeyman-master routine), that would mean something.
Yes - certification by a person or people is always better than certification by some anonymous system.
But that goes straight back to my usual condemnation of modern education systems. They prepare a person for academia, not for anything else.
Don't condemn all modern education systems and don't dismiss academe lightly. Some modern universities do an excellent job of turning out competent and well-educated people. In computing the top three in Britain are probably Oxford (which gave us Ted Codd - he learned relational algebra amongst much other mathematics as an undergraduate there), Cambridge gave us microcode, and Manchester invented pretty well all the early computing technology (both hardware and software). The US has some pretty good universities too, and some academics who have become major industry figueres (sticking to database, there's Michael Stonebraker, rather obviously); and so do a few other countries.
Tom
Viewing 15 posts - 616 through 630 (of 685 total)
You must be logged in to reply to this topic. Login to reply