Managing Large Application Portfolio with Technical Debt Related Measures
Jean-Louis Letouzey · 2016
Summary form only given. This presentation will show how some large organizations use some specific indexes and indicators of the SQALE method to manage the health of their software assets and optimize their maintenance budget. Major organizations have large legacy application portfolio. Some of these applications are old and their health is not very good. Any changes become costly and risky. This is due to the software entropy and the accumulation of technical debt. Refactoring of the complete portfolio is too expensive. A rational method is needed to prioritize refactoring activities. The method I have advocated and implemented in some very large banks, but also in software critical companies is as follows: · It starts with collecting two types of information about each application in the application portfolio: the relative business value and the annual maintenance level. · In addition, all of the applications are analyzed with a static analysis tool implementing the SQALE method like SonarQubeTM. This results in two indexes per application, the technical debt density and the business impact density calculated as defined by the method. · These two measures are visualized on a dedicated indicator called “debt map” which is the key piece of the prioritization logic. The presentation will provide a more precise description of these measures and graphical indicators and will present concrete examples from large corporations. The presentation will conclude with recommendations for the implementation of such an optimization program.