I've been charged with compiling an analysis of Cognos versus Microsoft
Analysis Services. Let me start by saying that I've personally worked
with both, and fully understand that each has its benefits and
drawbacks. However, I now need to draw up a comparison that shows the
conditions where one is preferable over the other based on
functionality, and I'm drawing a blank for the most part.

Could some people here give me a hand with this? I've read through some
of the threads & you guys seem to know what you're talking about. Any
help is sincerely appreciated!


E

--
Posted via http://dbforums.com

Eric

I would also be interested in the comparisons. I have never used Cognos and
have seen a quick demo at a computer show earlier in the year.

However, I have been told from two different sources, who after seeing
O2OLAP in action with Microsoft Analysis Services, that they can achieve
what they could through Cognos. The one source was a general ledger company
in the UK and the other a BI company in the business for 10 years.

Regards

Howard.Taylor@ domain below
www.o2olap.com


"ericdkauff" <member35486 (AT) dbforums (DOT) com> wrote

Please keep the abvious commercial plugs for your own product off this
newgroup. It's for real discussion not for promoting one's own product at
the expense of others. I work for Cognos, so I guess it's up to me to say
that we beat O2OLAP all the time... See how this goes? No one gets informed
and all posts are eyed with suspicion. Is you contribute with an
issues-related or technical content that is informative, great. Just don't
do it at the expense of other products and tools and don't nakedly promote
your own product... This is an ad-free medium...

"Howard Taylor [O2OLAP]" <Howard.Taylor (AT) o2olap (DOT) com> wrote

"ericdkauff" <member35486 (AT) dbforums (DOT) com> schrieb im Newsbeitrag
news:3226208.1060630562 (AT) dbforums (DOT) com...
Eric,

the question you state directs to a technical comparison of the data bases.
You haven't said anything about your sites' requirements, i.e. what kind of
users, reports, analysis, etc. Besides looking at the end user related
criteria you should have a look at the data base issues like data and user
volumes, etc. Please give more information and the people here will be able
and willing to help.

Kind regards,

Joerg

Hmmmm, I think I need to provide a little clarification here then:

First off, this is not a topic to advertise for either vendor. I simply
need some formal comparison points to use in future evaluations.

What I've been charged with is to perform a high-level comparison of the
two back-end technologies, insofar as to compare the strengths and
weaknesses in design & build capabilities and restrictions, storage
methods & considerations, and performance under various conditions (i.e.
do Cognos or Microsoft display performance differences when performing
certain calculations on the fly).

When I say back-end technologies, I'm simply talking about the
database/storage that each one uses. I've worked with Cognos quite a
bit, not worked with Analysis Services as much as I'd like, so I can't
put together a well-informed comparison. I'm well informed of all of
the front-ending capabilities for Cognos & Microsoft cubes, so I'm not
concerned with the actual analytics, just the design & storage.

Nigel, I keep up a bit with the OLAP Reports site, so I have a
tremendous amount of respect for your opinions and insights. I'm
talking about the Cognos PowerPlay suite, as I am mostly interested in
this from an analytics perspective, rather than
integrating/incorporating the planning and analytics frameworks. The
pieces I'm specifically interested in (rather than mentioning specific
products) are the PowerCubes, the Transformer modeling tool, and if you
want to include Enterprise Server, that's great too. I know there are
number of ancillary pieces in the Cognos suite (i.e. Access Manager),
but I consider those to be integral parts of the package.

Joerg, I don't have any site specific requirements. Tell you the
truth, we're a Cognos customer, but we work with both technology sets.
We simply don't have anything internally that specifically lays this
out for us.

Thanks for all of the follow-up. Hope to hear some constructive
feedback.


E

--
Posted via http://dbforums.com

The greatest deficit in COGNOS Transformer is that it is not able to
generate 'distinct count' measure in PowerCube!!! AS does provide such
measure calculation on the other hand.

But in terms of dimension modeling, COGNOS Transformer is much more powerful
than AS. (e.g. alternate paths drill-down, category grouping, etc. are not
supported in AS)

Both COGNOS PowerCubes and AS Cubes are highly compressed and portable.


ericdkauff <member35486 (AT) dbforums (DOT) com> wrote

Never had a problem doing this. There is a category count measure which is
a "distinct count" of categories. Considering each category should be
unique in the instance you describe, I don't see the issue?

ya

"Karaoke Prince" <karaoke_prince (AT) hotmail (DOT) com> wrote

ya,

Definitely, 'Category Count' measure solves this problem provided that
you have put the corresponding category as a dimension.

But in some circumstances, you can't do this. e.g. I would like to
count the number of customer/accounts in my credit card sales table
which has billions of transaction records with over 2 millions of
customers/accounts inside.

Obviously, I can't use category count in this scenario. Isn't it?


"ya" <john (AT) doe (DOT) com> wrote