OptiRun - A Platform for Optimized Test Execution in Distributed Environments

Abstract

Computer systems have grown to play an essential role in our society. Software testing is an important asset of the software development process, as it serves to evaluate the quality of the software object being tested. Some software tests can be automated to obtain benefits such as increased efficiency, expanded test coverage and saved time.

This thesis presents OptiRun; a platform for optimized execution of automated tests in distributed systems. OptiRun consists of two main elements; a controller, which is responsible for managing the distributed system as well as for test allocation, execution and result reporting, and a web-based user interface where users can upload and manage test scripts, request immediate test executions and schedule executions ahead of time as well as report failed test executions to the issue tracking system JIRA. Tests on the user’s level can be time-consuming, so the design and implementation of a test allocation mechanism named OptiX represented a major objective of this thesis. OptiX seeks to allocate tests to machines such that the overall execution time of test sets is attempted minimized. An alternative allocation mechanism was implemented for benchmarking purposes to better evaluate the performance of OptiX. An experimental evaluation and comparison of the two methods, using an extensive series of pseudo test sets, is also presented the thesis. This project is written in the Python programming language, and is built upon frameworks such as Selenium and Django.

OptiRun is intended to support the telecommunications company Altibox in the procedure of efficiently incorporating test automation as a practice in the testing process of their online web service TV Overalt.