Types of Impact

When you consider the tasks involved in upgrading software or applying change packages, you want to know about the impact to your existing system. There are various types of impacts, for example:

  • Structural impacts

    These could be impacts such as how a change to a field type or its properties could affect the structure of a record. Making changes to a field without considering its affect on the related record may cause runtime errors. Therefore, a developer analyzing the impact of a field change can prevent unforeseen errors during an upgrade.

  • Behavioral impacts

    These could be impacts such as changing a common library PeopleCode program called by other PeopleCode programs may affect the behavior of the calling program. In this case, Change Impact Analyzer would note that the calling PeopleCode is affected by the change to the library PeopleCode.

    Likewise, if a PeopleCode program writes to a field, the PeopleCode program impacts the field. Knowing this kind of relationship can help developers determine what other changes need to be made and help create a test strategy for the changes.

In typical PeopleSoft applications, there are a large number of relationships between definitions. Therefore, when Change Impact Analyzer performs an impact analysis, it generates a huge amount of data—more than what may be useful to work with. Consequently, it's important for you to determine the relevance of the relationships you want to analyze.

Change Impact Analyzer provides various features to make the impact searches more relevant. For example,

  • Selecting the rule set(s) to be used, which determines specific rules.

  • You can specify search paths, which cause a selective recursive analysis to be performed.

  • Executing an analysis for the impact of one definition on another allows you to manually explore specific relationships.

  • Using directed analysis, Change Impact Analyzer executes multiple search path analyses. When finished, it automatically produces a report, which can be filtered to allow further relevance.