Operational data store versus a data warehouse
What is the difference between an operational data store and a data warehouse?

    Requires Free Membership to View

    By submitting your registration information to SearchOracle.com you agree to receive email communications from TechTarget and TechTarget partners. We encourage you to read our Privacy Policy which contains important disclosures about how we collect and use your registration and other information. If you reside outside of the United States, by submitting this registration information you consent to having your personal data transferred to and processed in the United States. Your use of SearchOracle.com is governed by our Terms of Use. You may contact us at webmaster@TechTarget.com.

You will find many conflicting opinions on this. The following is the opinion shared by me and many of my consulting colleagues within The Data Warehouse Institute, of what the difference SHOULD be. Enjoy.

The operational data store lives in the operational support system environment. It typically serves the purpose of providing "near" real-time integration and reporting of data across disparate operational systems. It is designed for update. It is fed by operational support sytems, AND it will feed those systems. It is NON-historic. Many times operational applications get built upon the ODS structures. That ends the significant differences from a data warehouse. The following charactersitics are shared between an ODS and a DW. It is subject oriented, it is highly normalized. The data integration is enables using the same suite of ETL tools and EAI tools that enable the data warehousing environments.

The data warehousing environment lives seperate from the operational support systems environment. It serves the purpose of decision support, historical data mining, trendings, etc. It can be updated near real-time, but usually is updated on a premeditated scheduled frequency. It has architectural layers designed in 2 OR 3 tiers to support 3 roles: intake, distribution and access. It is designed for read only. It contains history. It is subject-oriented. The intake layer is normalized, the distribution layer introduces dimensionality and denormalization. The access layer consists of a suite of data marts designed for specific purposes (for trending analysis, etc), some relfecting star schemas others reflecting normalized schemas (for list management and reporting). It is loaded via ETL tools and EAI tools. It is typically accessed using BI tools.

For More Information


This was first published in July 2002

Join the conversationComment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.