First of all, thank you very much Nigel for helping me!!

Today I was told that you may drill down to detail in ProClarity and
Panorama with the drill through feature, when the cube is restricted to
browse up to some level in a dimension but there are still lower levels.
Is this ok??

The date volume is an issue which I knew was important but thought that
would bring no problem to us.

We have three different cubes planned (to be refreshed every day), but
we know modifications will come up. Let me describe the cubes we are
planning to use:

1. One of them has 5 dimensions and expects to receive 35.000 new
entries each day, wich must be kept for about two years.
2. Another one with 14 dimensions, and expects to receive 350.000 new
entries each day, which need to be kept for about five months.
3. Another one with 6 dimensions, and 20.000 entries each day, and which
need to be kept for about two year also.

Is this information useful enough to estimate the data volume? If it is,
what should it be?

I will try to answer your questions so you can help me further:

A. "What is the application?": The application is mainly to monitor and
analyse the daily operations of the company. Later on, if this
project is succesfull we will extend its application to staff areas.
B. "What platform(s) must it run on?": It must run on SQL 2000.
C. "Do you need local cubes for off-line use?": I would be very useful
to have local cubes.
D. "What skills do you have in-house?": Our in-house skills do not
include Analysis Services handling but they do know SQL very well, so
in case we choose AS we will need consultants to develop the cubes
and train our people so as not to depend on them for future projects.
E. "Do you need complex, fine-grained security?": I don't think we need
fine-grained security.
F. "Do you need Internet deployments?": Internet deployment is not a
must, though I recognise how useful it would be. But the thing is
that we don't want to spend much on licenses, so probably we are
analysing the possibility of some of us getting licenses and the rest
accesing the cubes via Excel.
G. "Do you need advanced variance reporting, such as bubble-up
exceptions or sophisticated alerts?": We could use advanced variance
reporting and sophisticated alerts.
H. "Do you need data write-back?": Write back could be useful.
I. "What sort of calculations -- (eg, do you need complex
multidimensional
calculations or currency conversions)?": I don't quite understand how
you define complex multidimensional calculations, but I can tell you
that we are used to invent new coefficients and ratios.
J. "Do you need highly formatted reports as well as an interactive
viewer?": We need to care about reports formatting, but I can't
define how "high" they are.
K. There is some budget restriction, but I can fight for it!... and
maybe get results.

Again Nigel, THANKS A LOT!!

Regards,

Bernard

--
Posted via http://dbforums.com

"BRiemann" <member35054 (AT) dbforums (DOT) com> wrote

Based on your answers, Analysis Services with either ProClarity or NovaView
sounds like an appropriate choice, especially if you are already committed
to SQL Server 2000. It can handle your data volumes without problems,
supports local cubes (unlike MicroStrategy), is faster than PowerPlay or
MicroStrategy and costs will be much lower. It also supports write-back,
unlike MicroStrategy or PowerPlay (though Cognos has other financial tools
that do provide write-back). Cube building in Analysis Services is easy
enough and usually takes less effort than preparing the data in the star or
snowflake schema. The security system is functional, but can cause
performance problems if you have a large number of roles to manage.

You will have to decide whether you prefer the ease of use of NovaView or
ProClarity; some people prefer one and some the other and it's really a
personal choice. Temtec Executive Viewer is also particularly fast and easy
to use, though less well known in the Analysis Services environment. Of the
three, NovaView is probably the most functional, supporting more complex
exception reporting, and Executive Viewer the easiest to use, having
particularly good support for printed reports, while ProClarity is a
competent all-rounder that is the most widely used third party front-end for
Analysis Services. Crystal Analysis Professional is also available, though
it's less good than these three.

Another option is to use a good Excel add-in as the client tool, such as
IntelligentApps, MIS Plain or XLCubed. These work well and are the best
choice for numerate end-users who are already familiar with Excel. You get
all the security and maintenance benefits of a fast OLAP server, but with
the familiarity and convenience of Excel. And they're cheaper than the
proprietary clients, but aren't suitable for Internet deployments. The OLAP
Survey 2 found that people using Excel add-ins with Analysis Services
reported the highest business success rates of all. NovaView and ProClarity
also have less functional, read-only Excel add-ins, designed for formatted
reporting rather than data entry.

You also have the option of using PivotTables in Excel 2000 or XP
(preferably the latter), though I would only advise this if budget is very
tight. They aren't very functional and don't perform well. But at least it's
an option for occasional users who can't justify a dedicated client tool.
You can also use Microsoft Office Web Components, which are an improved
version of Excel PivotTables. But don't be tempted by Microsoft Data
Analyzer as it's a very specialist tool and unsuitable for most users.

As I mentioned previously, you don't have to stick to any one of these, but
can use any combination you like, and there will undoubtedly be other
front-end options in the future from both Microsoft and third parties. And
hooking up a new client tool to an existing server app is easy.

Nigel Pendse
OLAP Solutions
http://www.olapreport.com

Bernard

I would like to invite you to an online demonstration of what is possible
within the AS environment using O2OLAP for Excel. From other posts on this
board you will notice that we encourage the use of the MSAS environment to
keep users information "OPEN". We also cater for numerous drilling options
including drillthrough to source data. Modelling, writeback and graphics are
built in. But, there are numerous other features. I believe our features
easily compete with the products mentioned so far in the thread.

Howard Taylor
Howard.Taylor (AT) o2olap (DOT) com
www.o2olap.com



"BRiemann" <member35054 (AT) dbforums (DOT) com> wrote