Metaheuristic based auto-scaling for microservices in cloud environment: a new container-aware application scheduling
Author:
- 2 June 2021
- journal article
- research article
- Published by Emerald Publishing in International Journal of Pervasive Computing and Communications
- Vol. 19 (1) , 74-96
Abstract
Purpose: The cloud is a network of servers to share computing resources to run applications and data storage that offers services in various flavours, namely, infrastructure as a service, platform as a service and software as a service. The containers in the cloud are defined as “standalone and self-contained units that package software and its dependencies together”. Similar to virtual machines, the virtualization method facilitates the resource on a specific server that could be used by numerous appliances.Design/methodology/approach: This study introduces a new Dragon Levy updated squirrel algorithm (DLU-SA) for container aware application scheduling. Furthermore, the solution of optimal resource allocation is attained via defining the objective function that considers certain criteria such as “total network distance (TND), system failure (SF), balanced cluster use (BC) and threshold distance (TD)”. Eventually, the supremacy of the presented model is confirmed over existing models in terms of cost and statistical analysis.Findings: On observing the outcomes, the total cost of an adopted model for Experimentation 1 has attained a lesser cost value, and it was 0.97%, 10.45% and 10.37% superior to traditional velocity updated grey wolf (VU-GWO), squirrel search algorithm (SSA) and dragonfly algorithm (DA) models, respectively, for mean case scenario. Especially, under best case scenario, the implemented model has revealed a minimal cost value of 761.95, whereas, the compared models such as whale random update assisted lion algorithm, VU-GWO, SSA and DA has revealed higher cost value of 761.98, 779.46, 766.62 and 766.51, respectively. Thus, the enhancement of the developed model has been validated over the existing works.Originality/value: This paper proposes a new DLU-SA for container aware application scheduling. This is the first work that uses the DLU-SA model for optimal container resource allocation by taking into consideration of certain constraints such as TND, SF, BC and TD.Keywords
This publication has 24 references indexed in Scilit:
- A novel nature-inspired algorithm for optimization: Squirrel search algorithmSwarm and Evolutionary Computation, 2019
- Application deployment using Microservice and Docker containers: Framework and optimizationJournal of Network and Computer Applications, 2018
- Genetic Algorithm for Multi-Objective Optimization of Container Allocation in Cloud ArchitectureJournal of Grid Computing, 2017
- Using dragonfly algorithm for optimization of orthotropic infinite plates with a quasi-triangular cut-outEuropean Journal of Mechanics - A/Solids, 2017
- SCARF: A container-based approach to cloud-scale digital forensic processingDigital Investigation, 2017
- Stochastic Resource Provisioning for Containerized Multi-Tier Web Services in CloudsIEEE Transactions on Parallel and Distributed Systems, 2016
- Preserving Privacy in Health Care Information: A Memetic ApproachJournal of Medical Imaging and Health Informatics, 2016
- Containerization and the PaaS CloudIEEE Cloud Computing, 2015
- 3D FEM comparison of lingual and labial orthodontics in en masse retractionProgress in Orthodontics, 2014
- Threshold prediction for segmenting tumour from brain MRI scansInternational Journal of Imaging Systems and Technology, 2014