Reducing instance acquisition lag to improve scaling out in the kubernetes cluster

Kanthimathinathan, Bharath Raj (2022) Reducing instance acquisition lag to improve scaling out in the kubernetes cluster. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Cloud infrastructure leverages containerization which allows developers to deploy application anywhere in a repeatable and consistent manner without library or package dependency failures. With kubernetes being the industry standard for container orchestration, there is still tremendous scope for improvement in kubernetes cluster autoscaling. Both vertical pod autoscaling and horizontal pod autoscaling is limited by the cluster node autoscaling which acts as a bottleneck. Hence, a new node autoscaling solution is required to overcome the limitation of cluster autoscaling. This paper proposes a kubernetes cluster autoscaling solution called ANA autoscaler which makes use of bash scripts which creates and adds a new cluster node in a dormant state, ready to use when needed during cluster scale out. Here, the instance acquisition time refers to the amount of time taken by the kubernetes cluster node to add to the cluster and make it usable. The results show 72% improvement in comparison with the reaction time taken for a node to be added in EKS amazon managed kubernetes cluster with a dynamic scaling setting. This will massively improve the scale out time and reduce the application performance degradation. This also indicates there is still scope for improvement in the reaction time of node autoscaling in kubernetes cluster.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Tamara Malone
Date Deposited:	16 Dec 2022 10:57
Last Modified:	16 Dec 2022 10:57
URI:	https://norma.ncirl.ie/id/eprint/5990

Actions (login required)

View Item