Software architecture

Software architecture
Antonystang · CC BY-SA 3.0 · source
Name	Software architecture
Field	Computer science, Software engineering
Introduced	1990s
Notable figures	Grady Booch, Mary Shaw, David Garlan, E. J. (Ben) White, B. W. Perry

Software architecture describes the high-level structures of a software system, the discipline of creating those structures, and the documentation that captures them. It connects stakeholder concerns, system organization, and project constraints to guide design, implementation, evolution, and governance.

Overview

Architecture situates a system within technical, organizational, and operational contexts such as those faced by IBM, Microsoft, Google, Amazon and Netflix. Practitioners draw on precedents from Xerox PARC, Bell Labs, MIT, Stanford University, and Carnegie Mellon University to address cross-cutting concerns encountered by teams in Silicon Valley, Bangalore, Shenzhen and government agencies like NASA. Historical milestones and influential works from figures at University of California, Berkeley and University of Illinois Urbana-Champaign feed into curricula at institutions such as Massachusetts Institute of Technology and University of Cambridge.

Architectural Styles and Patterns

Common styles include layered architectures used by Sun Microsystems platforms, client–server patterns exemplified by W3C systems, microservices popularized at Netflix and Amazon Web Services, service-oriented architectures explored by IBM and Oracle, event-driven patterns studied at Apache Software Foundation projects, and domain-driven designs influenced by practitioners associated with Eric Evans and Martin Fowler. Additional patterns arise from distributed systems research at Google Research, Microsoft Research and labs like Bell Labs, reflecting lessons from projects such as MapReduce and Hadoop.

Design Principles and Quality Attributes

Architects balance attributes such as scalability seen in Amazon retail systems, performance important to Intel hardware integrations, reliability demanded by CERN experiments, security concerns raised by incidents involving Equifax, and maintainability emphasized by academic programs at ETH Zurich and Princeton University. Principles like separation of concerns, modularity championed by Grady Booch and Edsger W. Dijkstra, and information hiding advanced at University of Toronto drive trade-offs that architects reconcile with stakeholder input from organizations such as World Bank or regulators like SEC.

Documentation and Modeling

Architectural documentation practices borrow notation and models from standards and groups such as Object Management Group (UML), viewpoints promoted by IEEE standards, and frameworks like The Open Group's TOGAF. Modeling tools and repositories developed by companies including Atlassian, Sparx Systems and research from Carnegie Mellon University enable diagrams, component descriptions, and rationale capture used in projects at Siemens and General Electric. Recording decisions with techniques influenced by practitioners at Microsoft Research and historians at Stanford University supports traceability for auditors like PwC and KPMG.

Evaluation and Analysis

Evaluation methods draw on case studies from NASA missions, benchmarks from SPEC, and experiments reported by ACM and IEEE conferences. Techniques such as scenario-based analysis and architecture tradeoff analysis methods were formalized by researchers affiliated with Carnegie Mellon University and University of California, Irvine, and are applied in safety-critical domains overseen by agencies like FAA and European Union Agency for Railways.

Implementation and Deployment Considerations

Implementations must integrate toolchains from vendors like Red Hat, Canonical and Microsoft Azure, and runtime platforms such as Kubernetes ecosystems or Docker containers used at Google and Amazon Web Services. Continuous integration and delivery practices traced to engineering cultures at Facebook and ThoughtWorks affect deployment pipelines, while constraints from cloud providers like Microsoft Azure or regulatory environments in regions like European Union influence operational architecture.

Governance, Process, and Organizational Context

Governance spans decisions made by boards and teams in corporations such as IBM, Accenture, and Deloitte, and is shaped by methodologies promoted by Scrum Alliance and standards bodies including ISO and IEEE. Organizational structures from Conway’s insights linked to Melvin Conway and scaled patterns used at Spotify and ING Group demonstrate how team boundaries, procurement policies in agencies like U.S. DoD and strategic roadmaps at firms like SAP steer architectural outcomes.

Category:Software engineering