An Evaluation of Service Mesh Frameworks for Edge Systems

Elkhatib, Y. and Povedano Poyato, J. (2023) An Evaluation of Service Mesh Frameworks for Edge Systems. In: The 6th International Workshop on Edge Systems, Analytics and Networking (EdgeSys 2023), Rome, Italy, 08 May 2023, ISBN 9798400700828 (doi: 10.1145/3578354.3592867)

Abstract

Service Mesh Technologies (SMTs) are increasingly popular in simplifying the networking between microservices. They allow one to declaratively and programmatically define service-to-service policies and interactions, and take all sorts of network management logic (e.g., traffic splitting, request tracing, security, reliability) out of the application. This simplifies the development of microservice architectures, which are widely used in cloud and edge applications. However, the suitability for different SMTs for use in edge applications is unclear. Thus, this work compares the two most popular SMTs (Istio and Linkerd) in terms of performance and overhead for resource-constrained devices. Through extensive experimentation and comparing with a baseline of standard networking in a Kubernetes cluster, we identify that Linkerd offers a more edge-friendly SMT option in contrast to Istio. Overall, Istio's communications are ≈10% slower than Linkerd at an increased 1.2--1.4x more memory and ≈1.2x more CPU utilization.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Elkhatib, Dr Yehia and Povedano Poyato, Mr Jose Luis
• Authors: Elkhatib, Y., and Povedano Poyato, J.
• College/School: College of Science and Engineering > School of Computing Science
• ISBN: 9798400700828
• Copyright Holders: Copyright © 2023 Copyright held by the owner/author(s)
• First Published: First published in EdgeSys '23: Proceedings of the 6th International Workshop on Edge Systems, Analytics and Networking
• Publisher Policy: Reproduced under a Creative Commons licence

Project Code	Award No	Project Name	Principal Investigator	Funder's Name	Funder Ref	Lead Dept
315521		ABC: Adaptive Brokerage for the Cloud	Yehia Elkhatib	Engineering and Physical Sciences Research Council (EPSRC)	EP/R010889/2	Computing Science