A Decision Model for Selecting Serverless Platforms

Serverless computing has revolutionized cloud computing by abstracting the management of underlying infrastructure. This paradigm shift allows developers to concentrate on developing and optimizing application logic rather than managing the complexities of servers. Platforms such as AWS Lambda, Azure Functions, and OpenWhisk are gaining widespread adoption by enabling organizations to deploy code in response to events. However, selecting the right serverless platform becomes challenging for organizations as they must consider factors like scalability, integration capabilities, pricing models, and developer support. Moreover, decision-makers often lack specialized knowledge in every technical domain which requires them to continuously adapt to the rapidly evolving information and updates associated with these platforms. 

This study aims to assist decision-makers in selecting the most appropriate serverless platform based on their specific requirements. By developing a decision model, we seek to streamline the selection process and provide insights that significantly reduce the time and effort required to evaluate and compare available serverless platforms.

The replication package includes all the data utilized in developing the decision model:

Platforms: This sheet lists the platforms included in the decision model, validated by subject matter experts.

BFA Mapping: This sheet maps Boolean serverless features to corresponding platforms. A value of 1 indicates that the platform supports the feature, while 0 signifies that it does not.

NFA: This sheet details the mapping of non-Boolean features to serverless platforms.

S.F. Mapping: This sheet maps quality attributes to specific serverless features.

Feature Requirements: This sheet outlines the feature requirements expressed by case study participants for serverless platforms.

Case Study Demographics: This sheet provides demographic details of the case study participants, their current platforms, and the recommended platforms generated by the decision model ranked by percentage from highest to lowest.

All data included in the replication package were instrumental in achieving the study's objectives.