Recently I saw several PR issues regarding the sequential execution of repairing UTs.
It seems that some of our test scenarios will fail in certain orders. For one of the PRs I have already mastered the master[1]. There are also two PR[2][3] repair methods I think I need more people to review. [1] https://github.com/apache/incubator-dubbo/pull/2797 [2] https://github.com/apache/incubator-dubbo/pull/2807 [3] https://github.com/apache/incubator-dubbo/pull/2815
