Automating Stability and Power Management Testing for Server Platforms based on Xeon Processor

Abstract

Server Motherboards are the sole of Industry for powering the Internet and Cloud Computing. They are boards that act as a computing brain for the Data Centers. Manual testing of such complex data centers is time-consuming and a tedious job. The manual testing of complex server board is prominent to errors due to human intervention. Industry requires quick and accurate results. The most important part of testing is that the silicon chips designed with the platform architecture are compatible with the release builds. These release builds are a combination of firmware and the operating system which follows an iterative design-based approach. Every iteration has some modification and improvements over the previous one which has to be validated in a short period of time. This thesis presents the end-to-end automation testing structure and design for Server boards which has a rigid system with various control capabilities for hardware and support for multiple programming languages for validating the software packages. This implementation of this continuous testing approach helps to pinpoint the exact software and hardware build bugs at an early stage of design of the release con- figuration builds which reduce the overall throughput and costs for developing the Server boards.