We currently have a normalized group-wide data warehouse based on a UNIX/Oracle platform. To be able to get the data out to the users we want to create several data marts and are looking at using an ETL tool that ties in well with Oracle. We have looked at Oracle Warehouse Builder and although for simple ETL it seems it will handle the job, for more complex transformation logic it seems to suffer. My question is whether you have used the product successfully and whether you could recommend any other ETL products that may fit the bill.


Sid Adelman’s Answer: Take a look at Informatica, DataStage and Ab Initio. Give each one of them some of your more complex transformations and see how well they perform. The performance will include the vendors' difficulty in completing the transformations, your people's ability to understand how they did it and, if you have a large data warehouse, you will want to compare the ETL run times. Joyce Bischoff’s Answer: You may wish to look at AB Initio or Informatica products.

Chuck Kelley’s Answer: There are many good ETL tools on the market, including the one you spoke of – Oracle Warehouse Builder. Choosing one may not be so easy. You need to define the types of transformations that will most likely need to take place and then look at the products to see how they implement your needs. What do you use to build the data warehouse? Why not just use that ETL tool? I would be surprised if the ETL process for a data mart feed from the data warehouse would be more complex than building the data warehouse. In fact, I would expect it to be quite simpler!

David Marco ’s Answer: I would recommend looking at Ascential and Informatica. Both of these tools are high quality and work well with Oracle. At EWS we did a comparative study that you can find information on if you visit our Web site at www.EWSolutions.com.

Clay Rehm’s Answer: I have seen a demo of the product but I have not used it. There are a number of ETL tools that will work well with Oracle/UNIX. However, no matter what ETL product you are looking at, there are general guidelines you must follow before selecting and purchasing. Decide what the ETL tool must do, what is nice to have. Create a document and create categories, such as:

  • Performance
  • Meta Data
  • Extract
  • Transformation
  • Load
  • Integration
  • Security
  • Auditing, etc.

Within each category, list specific items the tool must be able to perform, such as load 20 million rows of data into a table within n hours. Once you have your list verified and approved by your sponsors, then it is time to narrow your list down and have vendors come in for a proof of concept. Based on the results of the proof of concept, select the vendor and tool based on how they meet your requirements.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access