Analytics

This part is dedicated to analyzing the architectural metapatterns, for if this book’s taxonomy of patterns is a step forward from the state of the art, it should bear fruits for us to pick.

I had no time to research every idea collected while the book was being written and its individual chapters published for feedback. A few of those pending topics, which may make additional chapters in the future, are listed below:

• Some architectural patterns (CQRS, Cache, Microservices, etc.) appear under multiple metapatterns. Each individual case makes a story of its own, teaching about both the needs of software systems and uses of metapatterns.
• There are different ways to split a component into subcomponents: Layers of Services differ from Layered Services. This should be investigated.
• An architectural quality may depend on the structure of the system. For example, a client request may be split into three subrequests to be processed sequentially or in parallel, depending on the topology (e.g. Pipeline, Orchestrated Services or Replicas), thus it is topology which defines latency. We can even draw formulas like:
  • L = L1 + L2 + L3 for Pipeline,
  • L = MAX(L1, L2, L3) for Orchestrated Services which run in parallel,
  • L = MIN(L1, L2, L3) for Replicas with Request Hedging.

Other smaller topics that I was able to look into made the following chapters:

Contents:

• Comparison of architectural patterns
  • Sharing functionality or data among services
  • Pipelines in architectural patterns
  • Dependency inversion in architectural patterns
  • Indirection in commands and queries
• Ambiguous patterns
• Architecture and product life cycle
• Real-world inspirations for architectural patterns
• The heart of software architecture
  • Cohesers and decouplers
  • Deconstructing patterns
  • Choose your own architecture

<< Mesh	^ Home ^	Comparison of architectural patterns >>