Legacy systems to cloud migration: A review from the architectural perspective

Abstract

Legacy systems are business-critical systems that hold the organization’s core business functions developed in a traditional way using monolith architecture and usually deployed on-premises. Through time, this system is exposed to improvement changes, increasing its size and number of functionalities, thus increasing its complexity, and maintaining it becomes a disadvantage to the organization. Migration to the cloud environment becomes the primary option to improve legacy application agility, maintainability, and flexibility. However, to take advantage of the cloud environment, monolith legacy application needs to be rearchitected as microservice architecture to fully benefit from cloud advantages. This paper aims to understand the motivation for cloud migration, investigate existing cloud migration frameworks, identify the target architecture for the cloud, and establish any empirical quality issues in cloud migration from the implementation point of view. To achieve those objectives, we conducted a systematic literature review (SLR) of 47 selected studies from the most relevant scientific digital libraries covering pre-migration, migration, and post-migration stages. The SLR outcome provided us with the primary motivation for the cloud migration, existing cloud migration frameworks, targeted migration architecture patterns, and migration challenges. The results also highlight areas where more research is needed and suggest future research in this field. Furthermore, our analysis shows that current migration approaches lack quality consideration, thus contributing to post-migration quality concerns.