The Effects of Parallelizing Builds in Continuous Integration Software

Abstract: Quick feedback in regards to build times is important in Continuous Integration. If builds become too long, it can hurt the rate of software development. There are multiple methods to reduce build times. One commonly suggested method is to parallelize builds. This thesis aims to investigate the effects of parallelizing builds in Continuous Integration software and provide support for whether parallelizing is a good way of reducing build times or not. We conducted an experiment consisting of running tests on different Continuous Integration software with different configurations. These configurations changed how many tests were executed and how many parallel build agents were used. The aspects that were observed and analyzed was how build time, average CPU usage and CPU time were affected. What we found was that parallelizing a Continuous Integration build drastically improves build time, while RAM usage and CPU time remains similar. This entails that there are no major consequences to parallelizing other than utilizing more threads and therefore using more of the available CPU resources.