We are using SonarQube 8.5.1 Developer Edition and we’ve configured our branch analyse to be done against master branch.
For each new feature that we implement we are creating a feature branch. From a feature branch we are creating subbranches for every task:
We are constantly rebasing our features branch with master, to get the latest changes that were released.
Recently quality gate started to fail for feature-1 branch. The branch was created 2 weeks ago, the quality gate was passing. After merging task-12 (for which the quality gate was passing) quality gate started to fail for feature-1 branch, it complains about the coverage, it is 73%, less than 80%.
I’ve created a new branch, feature-1-copy, which has the same content as feature-1 and the quality gate is passing for it (coverage is 81).
By comparing the analyse result for feature-1 with feature-1-copy I noticed that new lines to cover have different values: 396 for feature-1 and 687 for feature-1-copy.
Both branches are compared to master branch. Why sonar does not spot the same changes in both branches since they are identical and compared to master?