Keep Calm and Wait for the Spike! Insights on the Evolution of Amazon Services

Abstract

Web services are black box dependency magnets. Hence, studying how they evolve is both important and challenging. In this paper, we focus on one of the most successful stories of the serviceoriented paradigm in industry, i.e., the Amazon services. We perform a principled empirical study, that detects evolution patterns and regularities, based on Lehman's laws of software evolution. Our findings indicate that service evolution comes with spikes of change, followed by calm periods where the service is internally enhanced. Although spikes come with unpredictable volume, developers can count in the near certainty of the calm periods following them to allow their absorption. As deletions rarely occur, both the complexity and the exported functionality of a service increase over time (in fact, predictably). Based on the above findings, we provide recommendations that can be used by the developers of Web service applications for service selection and application maintenance.

The athletic heart syndrome in web service evolution

“…Typically, these services have been around for long and most likely they do not follow the most recent trends, concerning the way that they have been designed and implemented. As we have already shown in prior studies, general‐purpose services have many responsibilities, 2 they have complex interfaces, and they grow bigger as they evolve 3 . Nevertheless, the developers of service‐oriented systems still have to use them because, as we said before, they solve problems that are frequently encountered in many different kinds of systems.…”

“…Like all software, general‐purpose services evolve over time. Their evolution may involve the expansion of the services with new functionalities to meet new requirements, or the maintenance of existing functionalities 3 . The evolution process is under the control of the service providers.…”

“…In our prior efforts, 3 we investigated the issue of service evolution from a theoretical perspective. Specifically, we studied a set of popular, long‐lived, general‐purpose Amazon services that cover different domains, to see if their evolution follows Lehman's laws of software evolution 17 .…”

“…Our main research objective is to provide the developer with a pattern that formally specifies the key criteria for selecting a general‐purpose service that evolves according to his expectations . To identify these criteria and specify them in the form of a pattern, we perform a more detailed study of the services that we investigated in our previous work 3 . As we already know that these services live normal lives that respect Lehman's theory of software evolution, our goal is to identify the properties that indicate proper service evolution.…”

An integration-oriented ontology to govern evolution in Big Data ecosystems

“…They provide a catalog of changes, however they do not provide any approach to systematically deal with them. Other similar works, such as [28], empirically study API evolution aiming to detect its healthiness. If we look for approaches that automatically deal with such evolution, we must shift the focus to the area of database schemas, which are mostly focused on relational databases [24,17].…”