Impact Analysis

This is a suggested and typical approach to performing an Impact Analysis. Impact Analysis is beneficial when there is a need to analyze the consequences of changing a field in a Copybook.

The following can be customized or tailored by the user to meet specific, individual requirements:

STEP ONE:

• Use the Data Dictionary to identify and narrow down a list of possible seed fields.

  • Example: Search for variables having a specific token as part of their name, “ACCT-NO”, “ACCTNO”, “ACCTNUM” 

  • Example: Search for variables of a specific length and/or data type

• Capture the list of seed fields, along with application components where they are defined (Copybook, Program, etc.).

• Validate with a subject matter expert (SME) for duplicates or any ambiguity.

• Capture final list of seed fields, their attributes and details of Program/Copybooks where they are defined.

STEP TWO:

• Identify all application components directly or indirectly impacted by seed fields identified earlier. For each Program or Copybook identified in step one, use the Connections report to capture additional application components impacted.

  • Example: A Program will be referred to by a JCL, has the ability to call sub-programs, refer to data sources like files, and include other Copybooks. All of these components require a recompile or remediation, as needed, so this becomes a useful input for further estimation.

  • Example: Leverage the Data Source Structures Map report to further understand if the seed variables are part of a Data Source Structure or record layout. The details of these data sources—like the record layout— can be identified upfront for preparation of test data.

STEP THREE:

• Capture lines of code in Programs where seed fields are referenced. Use Variable Impact to identify lines of code where the seed fields are referenced. The Statement Collector report can also be leveraged for the same requirement, as it also captures the type of statement where the variable is referenced, in addition to IF/ELSE blocks if the variable is referenced inside a conditional block.

  • Example: The variable of interest can be a part of an assignment (MOVE), conditional (IF/ELSE, Evaluate), String Manipulation, UI statements, etc. Some statements would impact additional variables—like in a MOVE—or have no impact on any other variable, such as a DISPLAY statement. If the variable of interest is part of a CALL statement to another program, then the called program can be considered for further analysis.

STEP FOUR:

• Identify possible scenarios leading to statements involving seed variables or variables of interest. Use the Business Rules report to not only understand the actual lines of code that are impacted, but also the possible scenarios and conditions that lead to the variable of interest. This rule set or rule stack—which includes a sequence of conditions leading to a line of interest—accelerates creation of test cases and test data to post the changes for Impact Analysis.