Serverless Computing in Enterprise Application Integration : An Organizational Cost Perspective
Leung, Javier (2021)
Leung, Javier
2021
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2021052110253
https://urn.fi/URN:NBN:fi:amk-2021052110253
Tiivistelmä
Serverless computing has become increasingly popular as a deployment model for microservices. However, not much has been written about its applications in enterprise application integration (EAI), namely the practice of sharing data, business rules and processes between independently designed applications within an organization. Moreover, there is yet little consensus on what constitutes an optimal EAI architecture, making it difficult to evaluate the efficacy of serverless in its context.
Using the design science methodology, this research aims to (i) develop an abstract model for understanding EAI architecture, and (ii) evaluate the efficacy of serverless computing in EAI based on this model. Drawing from research in agile software development, organizational structure (U-form and M-form organizations) and transaction cost economics (firm boundaries, markets), we establish a theoretical framework in which EAI architectures can be understood as vertical hierarchies and horizontal networks involving tradeoffs between technological and organizational factors.
From this framework, we first define an abstract model of EAI architecture as a structure that seeks to economize on the production, coordination, and vulnerability costs of organization, while minimizing transaction costs incurred from search, decision-making, and enforcement. Second, we describe the main EAI architectural patterns in terms of our abstract model and examine how serverless emerges from this context. Lastly, using a hypothetical business case in which several information systems applications must be integrated behind a customer-facing channel, we design and implement a serverless integration runtime using Amazon Web Services and evaluate its efficacy as an integration architecture based on our model.
Using the design science methodology, this research aims to (i) develop an abstract model for understanding EAI architecture, and (ii) evaluate the efficacy of serverless computing in EAI based on this model. Drawing from research in agile software development, organizational structure (U-form and M-form organizations) and transaction cost economics (firm boundaries, markets), we establish a theoretical framework in which EAI architectures can be understood as vertical hierarchies and horizontal networks involving tradeoffs between technological and organizational factors.
From this framework, we first define an abstract model of EAI architecture as a structure that seeks to economize on the production, coordination, and vulnerability costs of organization, while minimizing transaction costs incurred from search, decision-making, and enforcement. Second, we describe the main EAI architectural patterns in terms of our abstract model and examine how serverless emerges from this context. Lastly, using a hypothetical business case in which several information systems applications must be integrated behind a customer-facing channel, we design and implement a serverless integration runtime using Amazon Web Services and evaluate its efficacy as an integration architecture based on our model.