Gadhe, Sunil Suresh (2024) Dynamic Infrastructure Scaling Mechanism in Preallocated Resource Environments: A Practical Deployment Approach with Apache Solr. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (805kB) | Preview |
Abstract
In today’s world, it’s crucial to optimize both the design and deployment strategies for applications. Traditional approach to scale out infrastructures on the cloud involves the use of the host metrics which include CPU, Disc I/O and memory usage. New nodes are brought, or nodes are removed from a node cluster based on given metrics conditions being met. However, this approach is ineffective in environments with preallocated resources, such as an Apache Solr cluster. In these configurations, a certain amount of memory is dedicated to the Solr service that implies different and constant memory usage at various queries requests. This can lead to a negative impact on the functioning of the application and results to time delay. To maintain optimal application performance, manual intervention is necessary to add and remove nodes from the cluster. Due to very high numbers of query requests, then the cluster must have maximum nodes in order to control the numerous Query requests. However, this approach results in resource wastage when the number of queries is relatively low. This paper introduces a dynamic scaling algorithm and deployment strategy that leverages a scheduler machine to address the challenge of efficient resource utilization in Apache Solr clusters. The overall goal of the study is to improve resource usage, extend the Solr cluster sustainably, avoid waste, and ensure the application’s high throughput during the busiest period, where some focuses are associated with each goal. The dynamic scaling algorithm will follow chosen metrics like the rate of Solr queries and the latency over the time frame of 1 min, or 5 mins. It will then scale in/out the size of the Apache Solr cluster to the production level in the most efficient manner possible.
Item Type: | Thesis (Masters) |
---|---|
Supervisors: | Name Email Gupta, Shaguna UNSPECIFIED |
Uncontrolled Keywords: | Cloud computing; Dynamic resource allocation; Scheduler machine; optimal resource allocation strategy; Service Principle; Azure Cloud; Algorithm for Resource Allocation; Resource scheduling; Resource management; VMSS; Apache Solr |
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: | Ciara O'Brien |
Date Deposited: | 03 Jul 2025 10:10 |
Last Modified: | 03 Jul 2025 10:10 |
URI: | https://norma.ncirl.ie/id/eprint/8014 |
Actions (login required)
![]() |
View Item |