SEAByTE: A Self-adaptive Micro-service System Artifact for Automating A/B Testing

TL;DR

SEAByTE is a framework designed to automate and improve the efficiency of continuous A/B testing in micro-service systems through self-adaptive principles, enabling more autonomous experimentation processes.

Contribution

It introduces a novel artifact that automates the experimentation pipeline for micro-services using self-adaptation, addressing the need for more automated A/B testing.

Findings

Demonstrates SEAByTE's effectiveness in automating A/B testing processes.

Shows improved decision-making speed in micro-service evolution.

Provides a concrete example of self-adaptive experimentation in practice.

Abstract

Micro-services are a common architectural approach to software development today. An indispensable tool for evolving micro-service systems is A/B testing. In A/B testing, two variants, A and B, are applied in an experimental setting. By measuring the outcome of an evaluation criterion, developers can make evidence-based decisions to guide the evolution of their software. Recent studies highlight the need for enhancing the automation when such experiments are conducted in iterations. To that end, we contribute a novel artifact that aims at enhancing the automation of an experimentation pipeline of a micro-service system relying on the principles of self-adaptation. Concretely, we propose SEAByTE, an experimental framework for testing novel self-adaptation solutions to enhance the automation of continuous A/B testing of a micro-service based system. We illustrate the use of the SEAByTE artifact with a concrete example.