DWS Basic Questions

[Accpac2010]

Hi There,
I am considering creating a reporting database for our financial group to allow them to slice and dice data from various views. This database will combine data from HR, finance and operation. My question is the type of analysis needed is down to the Line details level.
Does this mean that I have to feed this database with the daily transaction data from all systems (Fianance, HR, operation,etc)?
I don't know what column in these tables they will be interested in, do I then need to copy same table structure from the transaction databases to the new reporting database?
Now assuming that I imported transaction table for 5 or 6 different system into the reporting database, do I now need to create a new datawarehouse database to create fact and dimension table and be optimized for reporting?
Thanks a lot for your feedback
Bruce

 

[blom0344]

Bruce, whole libraries are written on this subject , but this is usually condensed to the Kimball vs Inmon theories.
Now, best of both worlds could be:

>>>>> Datavault <<<<<<

http://danlinstedt.com/

Ties Blom

 

[RiverGuy]

If you use the dimensional modeling methodology, which is the Kimball method blom0344 described above (and one of the most popular data warehousing methodologies), then generally, you will want model data at the lowest level of granularity possible. Your data warehouse structure will not be a copy of your transactional system. Your financial, HR and operational measures won't even be in the same table. You will need to meet with your users and business experts to decide which data to include. I suggest you take a look at the book The Data Warehouse Toolkit by Ralph Kimball if you decide to go with dimensional modeling.

 

[johnherman]

You generally scale the basic fact tables to the lowest level of granularity that will be required today, or in the future. Then you can build summary tables (also called data marts) based on the requirements for reporting, drill down, export, performance, etc.

You may also need to conform dimensions. It's too deep a subject to get into here, but you need to have standard values and groupings for common business terms. If one system reports at the state level, but another at the zip code level, then the lowest common denominator would be "State".

-------------------------
The trouble with doing something right the first time is that nobody appreciates how difficult it was - Steven Wright

 

[sai22]

I have installed informtica power center 7.1.1 and oracle 9i on my windows xp machine.
I started the Informatica service and it starts fine.
Then I started Informatica repository service and it starts fine but Informatica server stops as and when repostiory server starts.

so when I try to execute workflow it throws the error 'cannot connect to the server'.

I checked the server configuration in workflow designer.Then I assigned the server to the worflow.When I reopened the 'Assign server window' none of the workflows were assigned to my server. It is not saving my settings.

I am stuck here. Please give me the solution.I need it as early as possible.

Your help is greatly appreciated.