-
Bug
-
Resolution: Unresolved
-
Major
-
None
I use recordIssues in a step following steps with dotnet build, dotnet test, dotnet publish.
It identifies 3 issues, 3 times, and correctly identifies them as duplicates that it collapses.
enabledForFailure is set, and the quality gate is configured as
qualityGates: [[threshold: 1, type: 'NEW', unstable: false]]
However, it consistently uses the last successful build as a reference build.
I used reset on the quality gate, expecting it to update the new state as the baseline.
The log shows that the reset is recognized and used, but it then uses the reference build anyway and fails the build.
[MSBuild] Repository miner is not configured, skipping repository mining [MSBuild] Reference build recorder is not configured [MSBuild] Obtaining reference build from same job (main) [MSBuild] Resetting reference build, ignoring quality gate result for one build [MSBuild] Using reference build '[redacted]/main #10' to compute new, fixed, and outstanding issues [MSBuild] Issues delta (vs. reference build): outstanding: 0, new: 3, fixed: 0 [MSBuild] Evaluating quality gates [MSBuild] -> FAILED - New (any severity): 3 - Quality Gate: 1 [MSBuild] -> Some quality gates have been missed: overall result is FAILED [MSBuild] Health report is disabled - skipping [MSBuild] Created analysis result for 3 issues (found 3 new issues, fixed 0 issues) [MSBuild] Attaching ResultAction with ID 'msbuild' to build '[redacted]/main #20'.
stage('MSBuild Gate') {
steps {
recordIssues(
sourceCodeEncoding: 'UTF-8',
enabledForFailure: true,
tools: [
msBuild(),
],
qualityGates: [[threshold: 1, type: 'NEW', unstable: false]]
)
}
}
In the past I’ve used reset quality gate in another job a while ago just fine.
