PLATFORMS: Dell PowerEdge 6100 (NT Server running VMARK's DataStage), HP 9000 (UNIX server running Informix database).
BACKGROUND: The Timken Company is a leading manufacturer of highly engineered bearings and alloy steels. Based in Canton, Ohio, Timken is a global company with recorded 1996 sales of about $2.5 billion.
PROBLEM SOLVED: Our legacy systems could not manage our need to consolidate data and generate reports in numerous formats for end users around the country. DataStage is being used to extract, integrate and cleanse data from our legacy systems then load it into the data warehouse.
FUNCTIONALITY: We have experienced outstanding results with DataStage. Some credit goes to the hardware, but DataStage is very well architected. The product is easy to learn, highly functional and extremely fast. We also like its self-contained programmability. The DataStage environment is point and click, similar to that of a flow-charting tool, which maps the data transformation process but with the ability to write code from within the tool.
STRENGTHS: It runs extremely fast, giving us the productivity boost needed to load the warehouse quickly, minimizing required downtime for our operational systems. With a file of 200,000 records-- approximately 80MB of data--we can separate records based upon field value in 90 seconds, identify unique records and sift out duplicates in five minutes, and load the whole database in 20 minutes. Writing directly to a flat file, the job takes less than five minutes. Now, the biggest bottlenecks in our system are caused by ODBC-calls from other applications. DataStage can handle extremely complex transformations. This means we never have to write code, like C programs, outside of the tool. It has its own embedded BASIC-like programming language which can be used to develop, test and document complex rules and transformations--all from within DataStage.
WEAKNESSES: The tool is very intuitive; however, it does take time for a new user to adapt to DataStage's visual paradigm.
SELECTION CRITERIA: Although we did evaluate several tools, we opted for transformation engine tools over code generators. Several factors ultimately led us to choose DataStage:
1) Performance: DataStage significantly outperformed the other products, executing the same routines in a fraction of the time.
2) Functionality: We could "join" flat files more easily using DataStage. With most of our data coming from mainframe flat files, this was a key differentiator.
3) Fit with requirements: We needed a tool to build a warehouse--that is DataStage's strength.
4) Price: DataStage is an incredible value compared to other tools on the market. It is much less expensive than other products.
DELIVERABLES: DataStage loads our warehouse from a legacy VSAM mainframe. We may also use it to build data marts. Right now we are using a 4GL to populate the data marts from the warehouse, and we will probably use DataStage to incrementally update the warehouse and data marts simultaneously.
VENDOR SUPPORT: Relational Solutions, a Cleveland-based systems integrator specializing in data warehousing, helped us set up DataStage. Vendor support has been very good. The technical support has been fairly responsive. I have never talked to a live person on an initial call, but I leave a message and they always get back to me within a reasonable amount of time.
DOCUMENTATION: The documentation is on-line and easy to follow. It has helped me to solve several problems without calling tech support.
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
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access