Zoom In, Zoom Out in the Complex IS

 

• Zoom In, Zoom Out in the Complex IS

• Why a “high level” <>  “granular” mapping of an Information System?

• At a time of accelerating digital transformation, massive movements towards the Cloud, and the surge of "as a Service", it seems useful to be able to understand in general, or in detail, how an Information  System Information works. IT architectures are often complex, often still partially or largely on prem'. And even though they are essentially in the Cloud, they tend to continually become more complex (and expensive ☹). This is the phenomenon of “entropy” very much at work in IT. We propose to dynamically deliver  a map of an information system, from a “high level” vision to a granular vision, or vice versa, to allow everyone to access their needs.

• 2 technical biases:

• This mapping must be  dynamic , and reflect the exact reality of the system at time “t”. This mapping must be  technical, not declarative  : it will be based on real processes, that is to say the continuous analysis of flows (procedural code, ETL/ELT, ESB, transformation in the data visualization layer, etc.), and the analysis of certain logs to know the uses of the information.

• 1 - A “High Level” mapping of the IS

• This representation allows us to understand the importance of each “data point”:  where are the strategic zones in an IS , the “hot” zones, the “cold” zones, the main “information highways” between the large groups of the IF, and  conversely , the dependencies do not exist. Some statistics help IS architects to make an objective diagnosis of the state of their system, and to deploy action plans to reach their targets.

• 2 - First zoom: complete flows, and      uses of information

• This representation allows you to visually discover how the information has moved/transformed. By choosing 2 “data points” (table, fields, etc.) from the IS, potentially very distant, all intermediate transformations are highlighted.  This vision will allow data engineering teams to have a simple and factual view of complete flows in 2 mouse clicks, whether they are on their infrastructure, in the Cloud, or both.

• 3 - Second zoom: a granular representation       flows and their uses through data lineage

• This representation,  a real granular “data lineage”,  allows, starting from a “data point” (table, file, fields), to unfold the flows in all directions, whatever the data transformation technology, to know the sources or final targets (uses), the underlying code, the execution dates, and any useful technical details. The data engineer teams will be able to carry out impact analysis, resolve problems with chains that are slow or inoperative, too complex, etc.

• Thus, it is possible to have 3 types of mapping of an information system with variable granularity, and with bridges from one to the other.  This will allow different but complementary approaches to continually improve IS governance.

www.ellipsys-lab.com 

contact@ellipsys-lab.com