When XML report file is analyzed by the parsers, we're not expecting to get duplicate testcases and then a new occurence of a testcase with the same classname+name, Bamboo ignores the second occurrence. Now the problem comes when in some rear cases the classname /name appear in different testsuites.

Perhaps we could implement the parsers in such a way to prefix the test cases with the testsuite names so that we don`t ignore dupes from different testsuites