We have built 10 applications and libraries in our project. We did this in a shared ODC environment. The Code Quality powered by Mentor identified about 300 issues for our applications and libraries that we are currently working on. During the project, we received a dedicated ODC environment and migrated all of our applications there.
The very first Code Quality scan on this new environment showed barely 100 issues... even though we know there should be at least 300. With each new scan, 10 to 50 new issues seem to appear, but we still don't see all of them.
Can someone explain how the Code Quality scans are performed? Is everything scanned with every scan, or is it only a subset? Do we simply need to wait a few days for all the issues to appear?
The behavior you're observing with the Code Quality scans in your new ODC environment can be explained by how the AI Mentor System (AIMS) for Code Quality operates.
Analysis Frequency and Updates: The Code Analysis runs automatically every 12 hours. When migrating to a new ODC environment, the first scan doesn't include everything retroactively, but rather starts analyzing the latest versions of your applications and libraries that have changed since the migration. As a result, findings may take a few scans to accumulate and reflect the comprehensive results you are expecting. This is why you are seeing issues appear incrementally after every analysis How ODC Code quality works.
New Environment Context: Moving applications to a new dedicated ODC environment essentially starts with a "clean slate" for Code Quality scans. This means findings need to be re-detected, and any previously identified patterns in the shared environment's scans don't carry over automatically.
Missing Findings on Older Published Apps: Since newly introduced code patterns or issues may not be detected retroactively, some findings for older app versions might not appear in the Code Quality Console until specific applications or modules undergo a 1-Click Publish. Performing a 1-Click Publish ensures that the latest versions of your applications are included in the subsequent scans and all relevant patterns are detected How ODC Code quality works.
Subset vs Full Scan: Analyzing code for Code Quality isn't purely incremental or limited to subsets but depends on when an application was last published or modified. If no significant changes have been made since the migration and publishing processes haven't triggered scans for all your applications, some findings may take time to appear in the results.
By following the above steps, you should be able to surface all 300+ issues or patterns identified prior to migration over time. Note that resolved findings from your previous environment may not carry over and would need active confirmation in the new environment.