Workflow Engines in the Compute Continuum: a Comparative Analysis

The intersection of Cloud and Edge computing has ushered in a new era of distributed computing, known as Continuum Computing. This computational paradigm shift leverages the computational strength of Cloud infrastructure and the pervasiveness of Edge devices to handle massive workloads, reshaping the landscape of distributed computing. One of the most challenging applications of Continuum Computing is the orchestration of scientific workflows, known as Continuum workflows. This problem is tackled by Workflow Management Systems (WMSs). However, the variety of WMSs in the literature is based on different techniques due to the rapid evolution of these systems. This study aims to compare two WMSs, that are Pegasus WMS and OpenWolf, by analyzing their key orchestration strategies and evaluating the performance of the two systems on workflows designed for Artificial Intelligence (AI) tasks. This comparative analysis is the first step toward identifying the most suitable frameworks for deploying AI-driven pipelines across the Continuum.