When it comes to Data Warehousing, what are you really getting for your money?
These costs are typically budgeted – but the key is to make sure the Data Warehouse cost is not wasted.
First, we need to define “Data Warehouse”.
Some technical terms are all over the map. “Data Warehouse” is one of them, as definitions range from “All Enterprise Data” to “Only the Data used for Analysis or Reports”.
Let’s use the term “Warehousing Data” for this discussion. “Warehousing data” is the electronic storage of a large amount of data stored multiple times.
Is that your environment? Do you “warehouse” all your data?
Have you been holding off on creating a user-friendly Reporting/Analytics layer? Have you been told you need a Data Warehouse first before you can build a reporting/presentation layer? Here is an example diagram of what “Warehousing Data” might look like:
Consider these questions:
- How much does it cost each time the data lands?
- With the recent advances in hardware and Relational Database Management Systems (RDBMS), is landing the data, and warehousing it, still a good value?
- Is there an approach that would produce the business features required without multiple data landings?
- If each landing costs $250,000 plus a year and you remove one or two landings, how much does that save?
- Chances are it is the Presentation Layer data that returns the investment. What if we could keep that data fresh? Not just once a week or once a day but how about once an hour?
- Is keeping all data changes (Slowly Changing Dimension) a good value? What if we save changes per business requirements? What if the presentation layer could go back to a business point in time?
There are some organizations that need to warehouse all their data, but is that your requirement?
Answerport has developed a hybrid ETL utility set for SQL Server that allows you to improve your data ROI. Some of the features of this utility are:
- Data Lineage
- Administrative GUI for faster build and implementation
- Historical business point in time tracking
- High performance DML with parallel I/O page alignment
- High performance loading with parallel bulk inserts
- Parameterized configuration with enhanced operational readiness
If you are struggling with costly or timely reporting, we have the answer.
Please email me with any questions, specific topic ideas for future blogs or to share data warehousing experiences at: firstname.lastname@example.org.
About the author: Jim Chance is a Consulting Director at Answerport, a management and technology consulting firm. Jim has many years of experience in large scale database management and has assisted many companies with their reporting, data presentation and data management needs.