I have a pipeline which runs tests in several concurrent branches, using the "junit" step in each to gather test results. Aggregation of the test results works correctly, AFAICT, but from the test stability plugin, there are multiple items that show up on test result pages. Most of them say "No known failures." When a test has failed, one item indicates the failure, but most items still say "No known failures." It seems that though each test result is only gathered once, a history item is generated for each test every time the junit step is called. I assume this problem results from the assumption that the test stability history plugin would only be called once, which was fine in freestyle jobs, but not in pipelines. The problem doesn't break builds, but it is annoying and potentially confusing.