Integration Software

Metadata management as a key factor of success in Agile BI

Although Agile Development, and all of its variants, originates from and in the first place is intended for pure application development, it is also finding its way in BI. With theory and (good) practice still being far apart from each other, some key environment pre-conditions should be implemented upfont, supporting a sustainable Agile BI lifecycle. To start with, a corporate domain-specific logical DWH model provides your different scrum teams overall guidance and alignment throughout the different sprints. It’s a strategic BI masterplan in which your Agile BI development deliverables will fit in the next few years. This common plan also demands a common development environment allowing for the scrum teams to work efficiently and in parallel on your BI delivery towards business. Besides back-end and front-end BI development tools, metadata management is often underestimated as a tool or medium in order to support and speed up the agile development process.

The ability to manage metadata

…starts with making it available to everyone working in the different scrum teams. Metadata should be available for analysis and should be embedded in your development environment. The only thing you should be worried about as a BI developer, is the metadata questions you need to be asking yourself, not where you can find it nor how you should make it available or extract it. In other words, also for managing your metadata you need to have the right tools in place.

Impact analysis

An Agile way of working involves alignment and agreement on what you are working on between the scrum teams and individual developers. Changing existing code implies prudence. Knowing the impact of the changes you will be applying is key in successful release management. Missed impact, can lead to reports or dashboards not being able to execute anymore. Hence, metadata impact analysis should become a standard built-in step and it should cost you no time in order to find out everything about data lineage and much more. Another benefit of easy reportable metadata is that when it is versioned, you can compare different BI outputs (e.g.: semantic models, report definitions) before deploying or moving code, because you don’t want to forget the BI Support team doing bugfixing or quick changes independent from the Agile tracks.

CoMetReporter in support of IBM Cognos Agile Development

Embedded in the IBM Cognos environment itself, CoMetReporter provides ad hoc reporting on top of your IBM Cognos BI Framework Manager and Cognos Connection metadata to all developers, analysts and authors in need of metadata insights. In an agile context CoMetReporter is well-suited to provide fast and intuitive access to the metadata you require at development time. You can find more use cases as well as the technical details on CoMetReporter here.

Example of a metadata impact analysis report:

Metadata management as a key factor of success in Agile BI comet impact analysis

Leave a Reply



Twitter Feed

Contact us

three − = 2