We are now approaching the end of our first year (out of three) of CloudWave. In many ways, the first year of a research project the magnitude of CloudWave is always the most difficult. In this blog, I thought I would discuss my views on some of the aspects of first year project management, after having been in this position a good number of times now.
• Get all the pieces in place early. I’m a firm believer in having the whole end-to-end execution of the stack working in the first year, even if this execution only shows a narrow slice of breadth. Setting this as your goal forces convergence of the high level definition of the architecture, defining components and their external interfaces. Having your use cases running in year one will drive the establishment of your test bed infrastructure and set up your software engineering practices. You’ll also discover your mistakes early, such as incorrect assumptions about required physical resources, as well as set up the required intra-project working relationships needed for the success of the project.
• Focus in on your true goals. Between the time that the proposal was originally written and the end of the first year, a lot of time passes – probably at least two years. Things will have evolved – new technologies and standards will have burst into the market, and the strategic interests of industrial partners may have taken on a new direction. You may find that some parts of the original proposal require more emphasis than originally budgeted, and other parts are no longer well aligned with either the project, the marketplace, or the executing partner(s). Evaluate the results of the first year and think carefully about where you are going over the next two years. Focus on your core project goals and consider a DoW amendment if necessary to trim off any distracting tasks that are no longer consistent with helping you meet your project goals.
• Learn how best to get the project’s goals across. Take a critical look at how your project was presented in Y1 - did you effectively get the message across of what you want to achieve in the project and what you have achieved up to now? Reexamine partner exploitation plans and planned Open Source contributions to make sure that in both cases the project is on-track to make an impact.
So, where does CloudWave stand in light of these points?
We do have our architecture defined and are working on getting our use cases to run against the whole stack now. Some components will have to be refactored or expanded in functionality in year two but that’s okay – if we aimed for final definition of all functionality before starting to implement we would end up in an endless cycle of discussions – we believe that the framework architectural design is correct.
We put out of scope for year one a number of tasks that were originally planned in the DoW. I suspect that while some of them are more naturally started in year two, others should be pruned from the project.
We are interested in talking to other FP7 projects about collaborating on topics of joint interest. We will be hosting a webinar in November to talk about our approach to Cloud monitoring. I hope that you’ll give it a listen.