Continue in 2 seconds

Meta Impact, Part 2

Published
  • December 01 2005, 1:00am EST

Do you know the complete impact change has in your enterprise environment?

In the first part of this series, we examined some of the frequent fallacies typically seen at companies that do not regard meta data and a managed meta data environment (MME) as a necessary component in their overall enterprise data management strategy. These firms continue to manage their inventory of information systems through inefficient and redundant methods by not accepting the inherent benefits this vital knowledge resource provides to the enterprise environment. In this installment of this series, we will look at one of the most significant benefits that adoption of an MME can have to a firm by looking at impact analysis and the insight gleaned across the enterprise when making changes to applications.

Every action in the information enterprise has a consequence; some are positive, some negative. Yet many organizations make changes to their enterprise applications that run their business without having complete insight into the true impact and cost a modification has across the entire organization. In the past, information systems were often siloed and essentially independent from other applications in the enterprise. Digital exchange of data between applications was very limited and frequently done manually through data entry. Today's information enterprises have evolved into highly integrated compilations of applications that utilize a variety of data integration methods (e.g., batch, ETL, EII, messaging, XML, etc.) to quickly and efficiently exchange data between them. A data impact analysis is one method for assessing the true results of a change, not only to the specific application that originated the change request but also across the entire enterprise. It provides a means for reviewing the total impact on the enterprise from a data structure, application, environment, business rule and data movement perspective through use of the MME. This allows both business and technical users the ability to make informed decisions on what changes offer the best opportunity to improve efficiency in the enterprise. The analysis provides a very granular level of impact analysis down to the lowest level of granularity available through the MME. This analysis information is vital for making informed business decisions on expenditures of scarce corporate resources (time, people and cost).

The MME captures the key meta data about a company's IT applications and uses this meta data to generate an impact analysis. An impact analysis is a report that shows the effect of a potential change to an IT application not only locally but also across the enterprise (see Figure 1). In this example of impact analysis from the MME, the domain objects NAME and SALARY are illustrated. The analysis report shows that NAME is used in two operational applications and in the data warehouse environment. Further technical meta data, data steward, business rules and transformation processing meta data is available for analysis on this domain.


Figure 1: Sample Impact Analysis


To further understand the value an impact analysis from an MME can provide to your organization, look at the simple example in Figure 2. In this example enterprise, we have three operational applications and a data warehouse environment that contains a data warehouse and a data mart. Application 1 is the originating source for a CLIENT NAME in the enterprise. For our impact analysis example, let us say that because of continued global growth in the organization, the length of CLIENT NAME domain needs to be expanded from 50 to 60 characters to accommodate the increasing list of new clients. An organization without an MME could probably determine with some level of effort the impact this change would have on Application 1. The user interface, the application logic, database and operational reports of Application 1 would all require modifications to facilitate the change request. That gives us six change points (assuming three impacted operational reports exist). The data interfaces Application 1 has with Application 2 and 3 will need to be modified due to the change request. Total change points are now eight. Typically, organizations without an MME would stop looking for other areas affected by this change at this point, taking a very narrow view of the impact. In other organizations, an enterprise-wide analysis would need to be performed manually, which is redundant and requires rework each time a similar change occurs.

Figure 2: Example Enterprise


In a firm with an MME, the impact analysis immediately shows that Application 2 and 3 also need to be modified, because they also store CLIENT NAME. This adds four more affected points from each application, plus one change to a data interface between these applications. Total impact count is now 17. Next, we need to determine the impact to the data warehouse environment. For simplicity, let us say that only one extract, transform and load (ETL) process is affected by the change. Total impact count is now 18. The data warehouse database, the ETL process from the data warehouse to the data mart, the data mart database and the data mart reporting meta data layer are also affected by the change and require modification. That gives a total enterprise impact count of 22 points that require modification because of the proposed change request to Application 1. That is a 367 percent increase in impact points that are affected by this change request versus the initial limited perspective.

Would your firm find all of these changes? Without an MME, what methods would be utilized by your organization to determine the full impact of this type of change request? Organizations without an MME to perform this type of impact analysis are often left with incomplete views of changes that can potentially cause failures with downstream applications. Without an MME, your company is missing the opportunity to exploit this knowledge resource to improve operational efficiency and productivity, which, in the end, affects the bottom line.

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