Seminar: An Engineering Approach to DevOps in the Context of Software Defined Systems

Program

Abstract:

The complexity of the new generation of systems developed around 5G and IoT brings a set of important (critical) new challenges, both from a technical and a business perspective. To succeed with the development of these systems, organizations must have the agility to adapt to constantly evolving heterogeneous environments to deliver solutions faster, and solutions that are better adapted to user personal needs and environments to improve their everyday life. Over the last decades, the concept of Software-Defined System (SDS) has emerged in different application domains as a key enabling technology for the development of next generation systems. SDS allows decoupling software applications from the physical execution (hardware) platform on which they are deployed by introducing a software control layer that is responsible for the virtualization of the operating environment. Examples of Software-Defined Systems (SDS) include Software Defined Network (SDN) and Software Defined Data Center (SDDC). SDN is considered one of the most important building block of the new generation of systems. SDS brings many important potential benefits, including reduced cost (in terms of both equipment and operating costs), elimination of vendor lock-in, optimized resource usage, reduced footprint, reduced operation complexity, increased flexibility to adapt to different operating environment and user needs, and increased agility to take advantage of innovations to provide new capabilities as product differentiators, as for example advanced analytics and machine learning/AI capabilities.

An important consequence of the emergence of SDS is that software has replaced hardware as the main asset and product differentiator in this next generation of systems. This comes with important new challenges. Among other things, the success of an organization is now determined by its business and technical agility regarding software development. The success of a company directly dependents on its ability to beat its competitors at the software level, i.e. to develop, deploy, and operate new software capabilities faster and with better quality than its competitors.

Over the last decade, DevOps has emerged as a leading approach in the industry to increase developer productivity and system quality. It builds on the success of the agile approaches that have been used for software development, and aims at leveraging the agile philosophy from a global product perspective by integrating aspects like system operation and customer support together with development. However, much too often DevOps is simply implemented by combining agile development principles with a set of tools used to automate different tasks of the development, operation, and IT process. Not enough focus is placed on the definition, documentation, and systematic analysis of the overall process, and on tools to support the DevOps process itself. As a result, DevOps is often implemented and managed in an ad hoc manner, which makes process improvements arbitrary.

In this presentation, we will discuss: the origin and main principles of DevOps, main challenges related to the implementation and use of DevOps in the context of SDS, and how, from an engineering perspective, we can provide process, methods, and tools to support DevOps.

Short bio:

Francis Bordeleau is professor at École de Technologie Supérieure (ETS) in Montreal. He is also co-Founder and President of Cmind Inc and Chairman of the Eclipse Papyrus Industry Consortium (IC). He has over 25 years of experience in software and model-based engineering (MBE); researching, working, consulting, and collaborating with companies worldwide. Prior to joining ETS, he was Product Manager of SW Development at Ericsson (2013-2017), where his main areas of responsibilities included MBE and modeling tools. In this role, he was responsible for defining product specification and roadmap, developing business cases, managing budget, managing open source initiatives, managing research projects, and collaborating with other end-user organizations, suppliers, and research/academia. Prior to Ericsson, he was Founder and CEO of Zeligsoft (2003-2013); Director of Tooling Business for PrismTech (2010-2013); and Assistant Professor at the School of Computer Science of Carleton University (1997-2006).

Francis holds a B.Sc. in Mathematics from University of Montreal, a Bachelor of Computer Science from University of Quebec (UQO), and a Master in Computer Science and Ph.D. in Electrical Engineering from Carleton University.