Smart impact analysis runtime performance
Introduction
Several factors affect the runtime performance of the Smart Impact app. These factors also affect the performance of other apps and workflows that perform an impact analysis, for example:
-
The Code Guard, Cross System Impact Analysis, Impacted Custom Code and Release Assurance apps.
-
The following workflows from the Impact Analysis templates folder:
Factors affecting the time taken for an analysis
The following factors affect the time taken for an impact analysis to complete:
-
Whether the Smart Impact app runs in Express or Standard mode
Factors not significantly affecting the time taken for an analysis
The following factors don't significantly affect the time taken for an impact analysis to complete:
-
The size and number of Test Repositories used in the analysis (the Create Test Repository Cache workflow may be scheduled to run overnight).
-
The amount of data exported to Excel, and the number of worksheet and columns in an app or workflow’s Excel report.
Size of the change (number of changing objects)
The Smart Impact app is normally provided with a set of transports or ChaRM requests to analyze. One transport or ChaRM request may contain a small change (for example, one custom program or transaction). Alternatively, it may contain the objects required for a complete project or upgrade (for example, hundreds of programs, customizing table entries, classes and authorization roles).
More changing objects to analyze leads to a longer analysis time.
Size of the production system (number of used objects)
The Smart Impact app uses production usage data to assess regression risk. By default, the analysis is focused on the following executable object types. These object types are reported as most-at-risk by the Smart Impact Analysis app and some of the Impact Analysis workflow templates.
| Object type | Description |
|---|---|
| WAPA | BSP Application |
| IWSV | Gateway Service |
| PROG | Program |
| FUNC | Function |
| TCOD | Transaction |
A user with Administrator permissions may customize these defaults as follows:
-
Select the Administration > Configuration > Impact Analysis folder in the LiveCompare hierarchy.
-
In the TypesToFind section, deselect the object types to be excluded from the used, impacted and most-at-risk results. Note that if no items are selected, all the above object types will be used as the default value.
-
Click Save to save your changes.
These changes affect all subsequent runs of the Smart Impact app and some Impact Analysis workflows, but they don't affect any existing app or workflow results.
More executable object types to analyze leads to a longer analysis time.
Size of the development environment
LiveCompare maintains an object links cache database for the Smart Impact app’s Analysis System. It stores a set of object relations which is used to determine which executable objects are impacted by which changing objects.
This database is created and maintained by the Create Object Links Cache workflow, which should typically be scheduled to run once or twice each week. It is also updated by the impact analysis apps and workflows.
A larger development environment has more relations to analyze. This leads to a longer analysis time for the impact analysis apps and workflows, and a longer runtime when the Create Object Links Cache workflow generates an initial cache of objects.
Analysis parameters
The Smart Impact app has several parameters that affect the time it takes to complete.
Search depth
By default, the Smart Impact app and impact analysis templates search for referenced objects at a depth of 10, however an Administrator may specify a custom search depth as follows:
-
Select he Administration > Configuration > Impact Analysis folder in the LiveCompare hierarchy,
-
Set a value in the ImpactAnalysisDepth field.
-
Click Save.
Decreasing this value reduces the time taken for the analysis, however it may lead to less accurate results.
Compare ABAP?
The Compare ABAP? switch in the Smart Impact app’s Variant screen specifies whether most-at-risk objects will be compared on the specified Analysis and Comparison systems. Switch this off to reduce the time taken for the Smart Impact app to complete.
Compare Data?
The Compare Data? switch in the Smart Impact app’s Variant screen specifies whether the table keys for changing tables will be compared on the Analysis and Comparison systems. Switch this off to reduce the time taken for the Smart Impact app to complete.
App parameters may affect the runtime of the Smart Impact app, at the cost of less information and accuracy.
Maximum concurrent processes
By default, LiveCompare is able to run three workflows concurrently. This limit applies to workflows run from within an app, workflows run directly in the studio, and scheduled workflows. If this limit is reached, additional workflows are queued for execution until one of the running workflows completes.
A user with Administrator permissions may set the maximum number of concurrent processes as follows:
-
Select the Administration > Configuration > General folder in the LiveCompare hierarchy.
-
Set a value in the MaxConcurrentProcesses field (a value of 10 is recommended).
-
Click Save.
App workflows that are that are queued could give the perception of a long runtime.
Concurrent analyses using the same analysis system
You can run the apps and workflows listed here in parallel on the same analysis system. However, the Create Object Links Cache workflow places a write lock on the Analysis System’s Object Links Cache database. This means that impact analysis apps and workflows that are using the same analysis system will be paused until the Create Object Links Cache workflow finishes running.
If an analysis is shown as running but it is paused because its Object Links Cache database is locked, this could give the perception of a long runtime.