Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

October 8, 2002

Self Help

The OAMS: a data mart to support information technology operations

by James D. Newman

Continued from Page 1

To meet our functional and technical requirements, we concluded that a traditional data mart design could provide us with the flexibility the OAMS needed to add attributes and provide multiple reporting perspectives. We designed the database using dimensional modeling concepts containing fact tables and supporting dimension tables according to star-schema architecture principles.

The Abstract Model

In order to effectively capture information needed to track the impact of system outages, the OAMS had to define the various types of components that comprised an application and its interrelationships. We quickly found that there was no consistent structure among applications. A single application's configuration resembled a traditional hierarchical organization chart. However, when multiple applications were modeled, its structure resembled a complex arrangement of matrix and hierarchical organization charts. Logically, we designed the abstract model depicted in Figure 1 to capture the relationships between applications and their components.

Within the OAMS, components of systems are defined and then related. The terms in Table 1 were key to defining components in the OAMS.

Certain components can't be defined except in the context of their dependency to other components. For example, a platform instance can't be defined except in the context of the server on which it resides. In order to create components within the OAMS, certain other components must first be defined. The required dependency between components can then be established upon creation of the dependent component.

Facts and Dimensions

The abstract model comes to life through the design of the fact and dimension tables within the OAMS. One of the primary fact tables was built to represent the dependencies that make up an application. For instance, each record within the fact table could describe a relationship between application, application component, platform instance, and server. The fact table is a "factless" fact table; it doesn't contain any specific numeric metrics within the record.

Dimension tables support each of these abstract model components within the fact table. The dimension tables describe the characteristics and attributes for each component within the application like platform or cluster instances, and servers.

The dimension tables also contain information useful to many audiences. The operations group finds performance and change-history information. The business partners find contact information, status logs, and scheduled maintenance activities. The finance department utilizes asset management information, including purchase data, price, and depreciation schedules.

Rate This Article

Comments:

Optional e-mail address:

Future of the OAMS

The OAMS is a work in progress. Its implementation sets the foundation for enterprisewide configuration management solutions. The first iteration was focused primarily on collecting and displaying application dependency information for complex custom applications.

Future phases of the OAMS will incorporate additional layers of information about the IT environment. It will evolve significantly with the integration of system monitoring products and work management applications. Ultimately, the OAMS will capture other logical (nonphysical) asset components, such as application code.

During times of consolidation and austerity, a company must commit to capturing the intellectual capital of the enterprise. Organizations are now recognizing the need to capture the operational aspects of their critical applications since many of the individuals responsible for creating them are leaving the organization. Most data warehousing solutions are focused on marketing and sales or finance and accounting. The IT organization rarely benefits from having a data mart that supports its operational requirements. Applications such as the OAMS provide key operations-related information critical to an IT operations group and its business partners for now and for the future.

James D. Newman [jim.newman@acquitygroup.com] is a principal and founding member of Acquity Group, LLC where he leads the practice in business intelligence.