Microservice decomposition and message queues.

Maharjan, Rokin, 1994-

Microservice decomposition and message queues.

dc.contributor.advisor	Cerny, Tomas, 1979-
dc.creator	Maharjan, Rokin, 1994-
dc.creator.orcid	0009-0009-4372-1717
dc.date.accessioned	2024-07-17T14:03:57Z
dc.date.available	2024-07-17T14:03:57Z
dc.date.created	2023-08
dc.date.issued	2023-08
dc.date.submitted	August 2023
dc.date.updated	2024-07-17T14:03:58Z
dc.description.abstract	Microservice architecture has gained significant popularity in recent years due to its ability to improve the scalability, maintainability, and flexibility of software systems. In this thesis, we propose an approach for microservice decomposition using a Variational Autoencoder (VAE) based Graph Neural Network (GNN). We conduct a comparative analysis of our approach by evaluating its results alongside two other existing methods. For this purpose, we employ custom benchmarks that we developed in-house, as well as a widely recognized benchmark that encompasses both monolithic and microservice versions of the application. Furthermore, we address another crucial aspect of microservice architecture—efficient communication between services. Message queues play a vital role in facilitating asynchronous communication, enabling the development of scalable and resilient microservice systems. To identify the most suitable message queue for different scenarios, we conduct a comprehensive benchmarking study on four popular message queues: ActiveMQ Artemis, RabbitMQ, Redis, and Apache Kafka.
dc.format.mimetype	application/pdf
dc.identifier.uri
dc.identifier.uri	https://hdl.handle.net/2104/12811
dc.language.iso	English
dc.rights.accessrights	Worldwide access
dc.title	Microservice decomposition and message queues.
dc.type	Thesis
dc.type.material	text
thesis.degree.department	Baylor University. Dept. of Computer Science.
thesis.degree.grantor	Baylor University
thesis.degree.name	M.S.
thesis.degree.program	Computer Science
thesis.degree.school	Baylor University