Evaluating the effect of locking on multitenancy isolation for components of cloud-hosted services.

Ochei, Laud Charles; Ejiofor, Christopher Ifeanyichukwu

Abstract

Multitenancy isolation is a way of ensuring that the performance, stored data volume and access privileges required by one tenant and/or component does not affect other tenants and/or components. One of the conditions that can influence the varying degrees of isolation is when locking is enabled for a process or component that is being shared. Although the concept of locking has been extensively studied in database management, there is little or no research on how locking affects multitenancy isolation and its implications for optimizing the deployment of components of a cloud-hosted service in response to workload changes. This paper applies COMITRE (Component-based approach to Multitenancy Isolation through Request Re-routing) to evaluate the impact of enabling locking for a shared process or component of a cloud-hosted application. Results show that locking has a significant effect on the performance and resource consumption of tenants especially for operations that interact directly with the local file system of the platform used on the cloud infrastructure. We also present recommendations for achieving the required degree of multitenancy isolation when locking is enabled for three software processes: continuous integration, version control, and bug tracking.

Citation

OCHEI, L.C. and EJIOFOR, C.I. 2018. Evaluating the effect of locking on multitenancy isolation for components of cloud-hosted services. Advances in science, technology and engineering systems journal [online], 3(3), pages 92-99. Available from: https://doi.org/10.25046/aj030312

• Journal Article Type: Article
• Acceptance Date: May 11, 2018
• Online Publication Date: May 31, 2018
• Publication Date: Jun 30, 2018
• Deposit Date: Mar 11, 2019
• Publicly Available Date: Mar 11, 2019
• Journal: Advances in Science, Technology and Engineering Systems Journal
• Print ISSN: 2415-6698
• Electronic ISSN: 2415-6698
• Peer Reviewed: Peer Reviewed
• Volume: 3
• Issue: 3
• Pages: 92-99
• DOI: https://doi.org/10.25046/aj030312
• Keywords: Multitenacy; Tenant isolation; Locking; Cloud-hosted service; Cloud patterns; Software process; Bug tracking
• Public URL: https://rgu-repository.worktribe.com/output/228089