Automated Triage of Performance Change Points Using Time Series Analysis and Machine Learning

• DOI: 10.1145/3491204.3527486
• Publication Date: 2022-07-19T12:52:43Z
• Authors: André Bauer; Martin Straesser; Lukas Beierlieb; Maximilian Meissner; Samuel Kounev
• Keywords: DevOps

Performance regression testing is a foundation of modern DevOps processes and pipelines. Thus, the detection change points, i.e., updates or commits that cause significant in performance software, special importance. Typically, validating potential points relies on humans, which considerable bottleneck costs time effort. This work proposes solution to classify detect automatically. On test data set provided by MongoDB, our approach classifies with an AUC 95.8% accuracy 94.3%, whereas classification based previous current exhibits 92.0% 84.3%. In both cases, can save time-consuming costly human work.