Data Warehouse vs Transaction Log tables

Member 10095656

Good day, How could I go about justifying a Data Warehouse for the business when there are already detailed transaction log tables in the OLTP database. I would need to give examples of why a Data Warehouse would give more and better answers when there is already a history of all updates done on "Product" as an example. In other words. What do I put in the Data Warehouse which is not already in the transaction log tables. Maybe someone can provide some valuable answers.

CHill60

I think the first thing to point out would be ..."how do you intend to get that information back out of the transaction log tables in a timely and usable way?" There are many many resources on line that will present the argument for you e.g. Top 10 Benefits of a Data Warehouse | Datamation[^] amongst millions of others[^]

Member 10095656

Thank you for the reply and for the link .... Yes, I understand. One of the most popular arguments for having Data Warehouse is that it should enable the user to access data/information faster. I have been thinking about these points (Benefits of having a data Warehouse) for a long time now. At the moment we do not have a very large business or production database. The largest of tables transaction log) consist of around 8 million records with 40 fields. So, I don't think speed of access is our problem however, I do agree with some of the other benefits like: 2. Enhances Conformity And Quality Of Data 3. Boosts Efficiency "to have to gather data from multiple sources" In our case we have many different types of databases (data islands) most of them SQL Server... So, I think that the DW would solve the challenge of bringing everything together in order to give the user a better view of how the business when all data is brought together into one view ("to have to gather data from multiple sources"). Again, thank you for the reply and the information ....

CHill60

Quote:

Boosts Efficiency "to have to gather data from multiple sources"

It also lowers frustration! :) After I had posted my reply I had to spend (waste?) quite some time combining data from 3 different sources (oh - and we do have a DW here, it's just not comprehensive. So we've introduced a second one on a different platform - you couldn't make this up really) It's "doable", especially with modern connectors and tools, but a real PITA - I seem to waste more time sourcing data to then be merged, than I do actually doing the analysis to gain insight. Like I said - frustrating. Perhaps then that is your most compelling argument ... A DW makes it easier and quicker to bring real insight to the business, turn data into information, to drive down costs and raise profitability. (I'm starting to sound like an advert so I will stop there :laugh: )

Mycroft Holmes

One of the primary reasons for a DW is so reports and data analysis does not impact the production (relational database) system. I hope your "data islands" are compatible, mapping disparate systems to a single platform can be a stone cold bitch!

Never underestimate the power of human stupidity - RAH I'm old. I know stuff - JSOP