DevOpsCon June 2017

Runtime information of deployed software has been used by both business and operations units to make informed decisions under the umbrella term “analytics”. Especially performance information (e.g. execution times, throughput, CPU utilization) often comes in the form of time series graphs in dashboards or numbers in reports. I hypothesize that this is not the right abstraction to present performance information for software developers whose daily workflow involves producing high volumes of code in an IDE. I argue runtime information needs to be especially targeted for software developers, hence the title "Developer targeted Performance Analytics". Performance metrics should be intertwined with source code artifacts (e.g., method calls, loops) in the IDE to provide this information in the right context and aid the software development workflow. In this talk, I want to present outcomes of the core topic of my Ph.D., that manifests this approach as an Eclipse IDE plugin, PerformanceHat [1], an open source project that combines runtime performance information with source code and leverages machine learning to provide performance predictions for code changes. This should enable software developers to make data-driven decisions about their code changes. [1] http://sealuzh.github.io/PerformanceHat/