COG-310 Tips

adma76 · Post by **adma76** » Thu May 14, 2009 8:11 pm

Hi guys,
I would like to receive tips about TM1's developer exam (310), I hear from my co-workers, the exam is very ambiguous !!!

Thanks,
Aviad Malka
http://www.linkedin.com/in/aviadmalka

Alan Kirk · Post by **Alan Kirk** » Thu May 14, 2009 9:19 pm

adma76 wrote:Hi guys,
I would like to receive tips about TM1's developer exam (310), I hear from my co-workers, the exam is very ambiguous !!!

Wow, I didn't even know there WAS one.

I've always thought of us TM1 developer / admins as more the MacGyver type; we work by our wits, ingenuity and imagination rather than by regimented book learnin' and exams.

However yes, it's true; there is an IBM exam.

All IBM certifications:
http://www-03.ibm.com/certify/tests/test_index_bd.shtml

This one in particular:
http://www-03.ibm.com/certify/certs/47010601.shtml

However if the "Sample Test" contained on that page is anything to go by, "ambiguous" isn't the term I'd use. Pathetic may be. My jaw is still dropping from the first question.

1. An application developer needs to provide a unique name for the server containing the TM1 application. Which of the following tm1s.cfg parameters should be used?
A. DataBaseDirectory
B. AdminHost
C. ServerName
D. PortNumber

Duuuuuhhhh, maybe it's a trick question and it's really "PortNumber", yeah, that's it, they always start with a "D".

2. Why does an application developer create multiple hierarchies in a dimension?
A. To increase the number of dimensions and empty cells in a cube.
B. To increase the sparsity of the dimensions and empty cells in a cube.
C. To reduce the number of dimensions and empty cells in a cube.
D. To reduce the density of the dimensions and string cells in a cube.

Not so much "missing the point", as "couldn't hit it with a howitzer if it were pinned to the side of a barn". How about:

E. To support the needs of the underlying business which has specified what it needs. Then, and only then, do you look at how to fill those needs most efficiently. There will be occasions when you cannot in fact design a system in the most technically efficient way (and you just need to accept that) because doing so would fail to meet the aforementioned needs which is the gorram point of the exercise in the first place, Mr. Examiner!

3. When loading data into a cube, which of the following must be specified for each of the dimensions?
A. Consolidation
B. Simple element
C. Data variable
D. Attribute

Words, they fail me.

7. Which of the following must an application developer do to distribute formatted reports to users
through a web browser?
A. Save the report to a reporting folder.
B. Save the report to the data directory.
C. Save the report on the TM1 server.
D. Save the report to an application folder.

Just make sure not to ask "How do you easily hyperlink to a .pdf document in an Application folder from a web sheet", 'cos that answer would be REALLY embarrassing.
A: {Mutter, mutter} Oh look, an eagle!
B: Ah, er, well...
C: We might get around to uhm, implementing that, um, some day. Manana. Maybe.
D: Hey, why do you need to link to .pdfs anyway? Just put all of your content on web sheets!

The closest I've found to an ambiguous question was:

6. How many dimensions must a reporting cube and look-up cube have in common in order to share data?
A. 1
B. 3
C. 2
D. 0

The recommended answer is D which, yeah, technically true I suppose... but the ambiguous part is in relation to what they actually mean by a "look-up cube". My interpretation of a look-up cube is (for example) our zDates cube which is used to return information about any given period element. (What the corresponding period was last year, what it'll be next year, how many working days it has, blah blah.) That obviously doesn't "share data" with the reporting cube(s) at all; it merely supplements that data with its own data for the purpose of creating reports. I think that what is meant here is not so much a "look-up" cube as a "detailed" cube, with the "reporting cube" being a summarised cube.

(And while it's technically true that a summarised cube and a detailed cube don't need to share any dimensions... I can't help thinking that it would be a rare pair that shared NONE of them.)

lotsaram · Post by **lotsaram** » Fri May 15, 2009 1:34 am

I think Alan has pretty comprehensively covered it but from what I hear the "developer" certification exam is massively dumbed down, too easy, and pointless. There are better ways to spend $200 USD.

adma76 · Post by **adma76** » Fri May 15, 2009 5:46 am

FYI,
Each IBM's business partner need/Must a minimum pool of certificated developer and Analyst

Thanks,
Aviad Malka
http://www.linkedin.com/in/aviadmalka

Alan Kirk · Post by **Alan Kirk** » Fri May 15, 2009 6:32 am

adma76 wrote: Each IBM's business partner need/Must a minimum pool of certificated developer and Analyst

I don't doubt it. In fact, it gives me an idea.

New policy: If any of you want to be an official OLAPForums Partner (giving you the right to wear an official OLAPForums lapel pin and put "Official OLAPForums Partner" on your stationery), forward a US$200 cheque to each of the Admins (starting with me, of course). Unmarked, non-consecutive banknotes in a brown paper bag are also acceptable. (The more you send, the better your chances of passing.)

However you will have to take a rigorous test to ensure that you're qualified. This guarantees that we can maintain the high standards that people expect of official OLAPForums partners.

Here's an example of the test that you'll need to take to see whether you've earned your OLAPForums stripes:
1: OLAPForums is:
A: A Forum dealing with OLAP stuff
B: A World Of Warcraft portal
C: A hotdog stand in New York's Central Park
D: A World War I tank.

2: The "Post" button in the Post A Reply dialog is:
A: Used to buy stamps from the post office
B: A lump of wood sticking up out of the ground, possibly supporting a fence
C: Used to post a reply to a thread in the Forum
D: Something that is only displayed after noon

3: The scheduling of TM1 Chores in UTC rather than local time is
A: Annoying
B: Beloved by everyone, especially Alan Kirk
C: Infuriating, especially when you move on or off daylight savings time
D: You get a pass on this question if you choose A or C. You not only fail the entire exam if you choose B, you'll pay double for your next test.

4: The OLAPForums colour scheme is:
A: Tangerine with really cool "go faster" racing stripes
B: English country garden floral
C: WW II Luftwaffe camouflage, as used on ME 109s
D: Blue and white, but A or C would be kinda funky if harder to read.

Examiner's note to prospective examinees: The examination is worth only 25% of your pass mark. The cheques / cash are worth 75%.

BigDSter · Post by **BigDSter** » Fri May 15, 2009 8:26 am

Fantastic, cheered me up this morning.

Love the question about multiple hierarchies, just shocked us here

I had heard there was going to be official certification of TM1 Developer status at some point, but if it's based around this, I don't think I'll push too hard to get it ...

Alan Kirk · Post by **Alan Kirk** » Fri May 15, 2009 9:26 am

BigDSter wrote:Fantastic, cheered me up this morning.

We aim to please.

(Or considering a certain level of barbedness in the above posts, we sometimes please to aim.

)

BigDSter wrote:I had heard there was going to be official certification of TM1 Developer status at some point, but if it's based around this, I don't think I'll push too hard to get it ...

It would be unfair to single out Iboglix; speaking as someone who attended some Oracle Financials courses many years ago, and whose mind still boggles at the per-day price compared to the outputs gained, I don't think that there's a software company on Earth which doesn't milk the training and certification cow for all it's worth. (And "Partner / Official Distributor /Gold Reseller" (etc) requirements just add a bit more leverage to pry some extra cream loose.)

If there's a difference, it's in how hard they pull the teat and, to stretch an analogy, the quality of the fodder that they provide to the cattle. (That is, us.)

It's just that asking a TM1 developer what a server name is doesn't strike me as being first quality grain mixed with barley.

Marcus Scherer · Post by **Marcus Scherer** » Fri May 15, 2009 10:19 am

adma,
the questions are very non-ambiguous, even for me who didn't grow up with Shakespeare in school.
If you're earning your daily bread with rules and TI you will pass the exam without preparation. If you're used to advanced scripting in TI you may take a look at the basic parts in the manual before.

Regards, Marcus

Martin Ryan · Post by **Martin Ryan** » Sat May 16, 2009 8:33 am

Ambiguity comes down to personality type sometimes, I dislike multiple choices because I'm used to trying to see other people's point of view, so if I see an answer that could be right from a certain point of view, it confuses me.

7. Which of the following must an application developer do to distribute formatted reports to users
through a web browser?
A. Save the report to a reporting folder.
B. Save the report to the data directory.
C. Save the report on the TM1 server.
D. Save the report to an application folder.

I don't use web, so could be about to embarrass myself here, but distributing a report in normal TM1 uses the application folder, so D. But the application folder exists on the TM1 server, so C is also correct. I haven't actually investigated where the files end up getting saved, but I would guess they're somewhere in the Data directory, so yeah, maybe B. But it'll be in a sub folder in the Data directory, which you could term a reporting folder, so A's correct too.

If, however, you asked me how to distribute a report, I would say - go to the TM1 menu in Excel, select upload report to Tm1, update or create new application, go to the applications folder, choose the appropriate sub folder and save.

But then asking open ended questions would require the examiner to know what they're talking about and exams are about making money, not spending it on qualified examiners.

Martin

lotsaram · Post by **lotsaram** » Sun May 17, 2009 10:24 pm

Marcus Scherer wrote:If you're used to advanced scripting in TI you may take a look at the basic parts in the manual before.

It's just fantastic fantasy that the test (and apparently IBM) seem to assume that as a "TM1 developer" one will be writing all TI script via the wizard. This is a fairly deranged notion but I agree with Marcus, if you have been taught to ignore the wizard and set all variables to other then a quick review of what the wizard actually does is probably useful.

Note emphasis on what the wizard does, I'm sure someone like Mr Kirk could write volumes on what the wizard doesn't do.

Alan Kirk · Post by **Alan Kirk** » Sun May 17, 2009 10:33 pm

lotsaram wrote:
Marcus Scherer wrote:If you're used to advanced scripting in TI you may take a look at the basic parts in the manual before.
It's just fantastic fantasy that the test (and apparently IBM) seem to assume that as a "TM1 developer" one will be writing all TI script via the wizard. This is a fairly deranged notion but I agree with Marcus, if you have been taught to ignore the wizard and set all variables to other then a quick review of what the wizard actually does is probably useful.

Note emphasis on what the wizard does, I'm sure someone like Mr Kirk could write volumes on what the wizard doesn't do.

No, no; I'm good to go on what it does...

Driving you absolutely batty in trying to get from A to B by the least convoluted and time consuming path, that's one thing that it's superb at.

Making you utter a blood chilling "AAAAAARRRRRGGGHHH" followed by "WHY CAN'T i GET THIS THING TO DO X, Y AND Z???"; it's great at that too.

But its most powerful feature is...

It motivates you to learn how to write TI code properly.

I concur with you wholeheartedly; a real developer would never waste their time it Wizard-generated code. It's just far too limiting.

I think this is an uncomfortable sign that the test (or at least some parts of it) was set by someone who has read the manuals, worked through the tutorials, but doesn't fully understand how the application is deployed in the "real world".

John Hobson · Post by **John Hobson** » Mon May 18, 2009 7:11 am

I think it's interesting that Applix threatened to bring in on-line certification some time before the takeovers happened but they never managed to generate anything they dared impose on the resellers.

Microsoft too have been threatening to impose certification as a requirement for membership of the MS partner programme for 2 years now but guess what....?

This is a fairly deranged notion

Agreed I found the TI wizard useful in providing some very basic examples of valid TI code, but I have only ever used in anger to load simple data files on a temporary basis.

Perhaps we could all suggest questions for inclusion in the IBOGLIX test.

Here's my first couple:

Q: Your TI process is not terminating with the expected results. Do you

A. Go into debug mode and step through the process
B. Run the process and inspect the detailed execution log automatically created by TM1
C. Run in on a REALLY slow PC (or over a WAN) and watch the screens change
D: Add a whole load of otherwise unnecessary variables and asciioutput() commands to generate a rudimentary output log that gives you some idea why your script might have failed.

Q. You are looking at a screen showing the entire budget for you company. You are copying the company sales total to an email when you realise your finger slipped on the control key and you entered "C" not "Ctrl -C". Do you

A. Click on undo
B. Update your CV and start emailing it around
C. Pray your sys admin didn't use the TI wizard that automatically sets logging off and leaves it that way if the TI fails half way through. Now call him and grovel.
D. Try to remember all of the underlying data numbers.

rollo19 · Post by **rollo19** » Thu Jun 18, 2009 2:07 am

Enjoyed the entertaining run down thank you Mr Kirk - however it wasn't so funny when I sat the exam to be presented with questions akin to those in the sample and the one Mr Ryan quotes. But still - you'd have to know a reasonable amount to pass. I would suggest anyone sitting it covers everything in the training course, and more.

Mostly it seemed to be a test of semantics. I am sure the exam will evolve..

Wim Gielis · Post by **Wim Gielis** » Thu Jun 18, 2009 8:09 am

Hi

I took the test last Friday and it was rather easy if you're a day-to-day TM1 developer. I had 92% correct out of 63 questions.

You get 90 minutes to complete, which is way too much

Wim

Alan Kirk · Post by **Alan Kirk** » Thu Jun 18, 2009 8:15 am

Wim Gielis wrote: I took the test last Friday and it was rather easy if you're a day-to-day TM1 developer. I had 92% correct out of 63 questions.

You get 90 minutes to complete, which is way too much

No it isn't; they need to factor in nap time for the boredom that the questions induce.

paulsimon · Post by **paulsimon** » Thu Jun 18, 2009 9:13 am

Dear all

The test is not perfect - no test ever is.

For example, I would agree with Alan that the primary reason for multiple hierarchies is to meet business needs. However, the suggested answer that it also reduces the number of dimensions is true, albeit at the expense of losing the ability to cross-analyse. Whether that is an issue or not depends on whether the business has a need to cross-analyse. Eg if your have stores with alternate hierarchies by type and location, do you ever want to cross analyse Mega Stores by Location, or are you happy to analyse either by location or by store type? As ever it comes down to business requirements.

However, when I looked at the sample questions there are certainly some there that someone who knew nothing about TM1 development would not be able to answer. I wouldn't employ anyone solely on the basis of their having this certificate, however, I believe that it will help to weed out those who have only used TM1 for reporting, and have never done any development (I have interviewed a few of those people in the past), or those who have done some Essbase, etc, and assume that TM1 is the same.

Inevitably there will be some questions that experienced TM1 developers will not be able to answer. For example, I probably would not fair too well on Executive Viewer having only ever put together a few simple views, but then I have come across very few companies who have a license for this, so there has been little opportunity to gain exposure and little demand for it.

However, it does appear that the bias of the test is towards rules and TI which are the core knowledge that any TM1 developer should have.

Regards

Paul

Wim Gielis · Post by **Wim Gielis** » Thu Jun 18, 2009 9:43 am

Very true, Paul.

I can testify that a real TM1 end user will have much trouble with solving the test.

without really developing (rules, TI - even the wizard -, what to have in a dimension and which dimensions in a cube in order to satisfy business requirements, ...) it's a lot harder.

For instance, the were 2 questions on how to setup a drill process. Normal end users will not have done this yet I guess.

IMO the test is able to distinguish between end users and developers.

Greetz,

Wim

mspat936 · Post by **mspat936** » Wed Jun 24, 2009 3:12 pm

Does anyone happen to have any more of the exam questions/answers? I am a hyperion developer who needs to take the TM1-310 exam so my company can stay in compliance with IBM and keep our partnership. I have found a few questions but could use a better sampling of questions to prep for the exam. Thanks!

Steve Rowe · Post by **Steve Rowe** » Wed Jun 24, 2009 3:45 pm

Hi mspat,
Welcome to the forum!
I'm not sure that it would be appropiate for this thread to give any more detail away on what is actually in the exam. It may just be a tick in the box but it seems clear from the postings that anyone who is used to developing TM1 solutions would be able to pass the exam pretty easily.

Sorry this is not much help but I am sure you understand.
Cheers

jameswebber · Post by **jameswebber** » Thu Aug 15, 2013 5:42 am

My 2 Cents on the exam

I just read all the IBM manuals. I focused on TI and Rules as this was 50% of the exam.

You need 73% to pass and I only got around 75% from recollection.

Some of the questions are vague multi choice where I felt I understood the concept but the question and options were written in such a way I wasn't sure which one to choose.
Like Martin I perfer questions than multi choice.
I didn’t do as well in TI as I would have liked since a lot of the questions were about using the inbuilt TI mapping interface - I'm not sure who really does this.

I also should have studied up more on time considerations.

If you new to TM1, it would be worth considering the analyst exam first. I believe this would be easier but not as technical.

But I don't know if the cert if a great representation of my ability to code TI or write feeders.

TM1 Forum

COG-310 Tips

COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips

Re: COG-310 Tips