Container Scheduling in Kubernetes Clusters Using a Mixed Integer Linear Programming Approach

A large part of the services provided on the Internet resides in virtualized environments of "cloud" providers, whose IT infrastructures adopt, for the most part, the containers virtualization technique to house the various applications that must run independently and isolated. The managem...

Full description

Saved in:
Bibliographic Details
Main Authors: Martins, Camilla, De Souza Moura, Pedro Nuno, De Lucena, Sidney C., Vieira Campos, Carlos Alberto
Format: Conference Proceeding
Language:English
Subjects:
Approximation algorithms
Cloud computing
Clustering algorithms
Container Scheduling
Containers
Costs
Facility Location Problem
Kubernetes
Mixed integer linear programming
Scheduling algorithms
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A large part of the services provided on the Internet resides in virtualized environments of "cloud" providers, whose IT infrastructures adopt, for the most part, the containers virtualization technique to house the various applications that must run independently and isolated. The management of such containers predominantly adopts the Kubernetes solution, whose standard algorithm for allocating containers on physical machines (nodes) is not capable of optimizing the cost/performance ratio of the cloud environment. As a solution to this problem, this work proposes a mixed integer linear programming approach based on the Capacitated Facility Location Problem to optimize the distribution of containers in nodes of a Kubernetes cluster, taking into account the cost of installing a new node, the history of failures of the nodes in operation and the consumption of resources by each application running in a container. The experimental results show that the proposed approach significantly improves the performance of Kubernetes cluster concerning the conventional scheduling strategy while ensuring the efficient execution of applications demands.
ISSN:2157-023X
DOI:10.1109/ICUMT61075.2023.10333098