Trade-off during SA and LA

Hello everyone,

At the beggining of every phase in the Arcadia method, factors have to be defined to characterize the expected output of the layer.

In System need Analysis, those factors 'support or hinder the effective implementation of capabilities" (JL Voirin, Model-based System and Architecture engineering with the Arcadia method)

In Principle/logical Architecture, these factors ‘constraint an influence the architecture definition’

However, I have the feeling, through the examples given in the book and the exeperience on our own systems, that those factors are often quite the same : Performances, costs, life-cycle considerations (mainly maintenance) …

Does anyone has any feedback about that topic ? Is there a clear conceptual difference between those factors?

Hello Julien,

you are right, many factors influencing each perspective building are similar from one perspective to the other. However, their impact is different depending on the intent of each perspective.

For example, considerations on performance in System Need Analysis are about which performances are expected from the system in which context, what is likely to influence them in the system environment, etc., such as number of user requests required per minute, or variance of this number in different use cases.

In Logical and Physical Architecture Analysis, performance focus will be put on which system design elements, choices and alternatives can affect performance, such as request processing contents and technologies, computing resources capacity, etc.

And of course, different factors may appear specifically in any perspective. This is the case for reuse considerations for example, that will mainly influence solution perspectives.

Thank you for your quick answer Jean-Luc,

You say that " performance focus will be put on which system design elements,".
In the way I understand it, it means that those factors and focus have to be defined in parallel from the architectural construction, and not pre-defined.

I guess that one has to stay flexible when trying to define a ‘chronology’ for the steps defined for each perspective,

Rgds,