Scholarlink

<p>Cloud computing, with its significant potential for remote data storage and processing, has brought new computational advantages for distributed applications. The existence of bottlenecks and congestion in the cloud environment makes traffic management essential in a scalable environment. This paper introduces a complex solution using an optimized genetic algorithm to address scheduling and load-balancing challenges in cloud infrastructure. Additional parameters are integrated into the algorithm to assess resource status before scheduling, effectively preventing server overload or underload by strategically allocating tasks to processing servers. Furthermore, tasks assigned to heavily loaded or dense servers are seamlessly migrated through live migration of virtual machines to alternative servers, enhancing the overall load balance of the cloud infrastructure. The effectiveness of the proposed approach is evaluated through the CloudSim simulator and validated by deploying over a thousand virtual machines in the PlantLab dataset and Azure 2019. Simulation results show a significant improvement in service-level agreement execution compared to comparative methods. Additionally, a reduction in energy consumption has been observed, while the average number of virtual machine migrations shows significant improvement.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A new load-balancing approach for cloud computing based on an improved genetic algorithm

  • 期刊论文

  • 发表日期：

    2025年6月26日

查看全文
  • Samira Rezapourshahir,
  • Shahram Babaie

摘要

Cloud computing, with its significant potential for remote data storage and processing, has brought new computational advantages for distributed applications. The existence of bottlenecks and congestion in the cloud environment makes traffic management essential in a scalable environment. This paper introduces a complex solution using an optimized genetic algorithm to address scheduling and load-balancing challenges in cloud infrastructure. Additional parameters are integrated into the algorithm to assess resource status before scheduling, effectively preventing server overload or underload by strategically allocating tasks to processing servers. Furthermore, tasks assigned to heavily loaded or dense servers are seamlessly migrated through live migration of virtual machines to alternative servers, enhancing the overall load balance of the cloud infrastructure. The effectiveness of the proposed approach is evaluated through the CloudSim simulator and validated by deploying over a thousand virtual machines in the PlantLab dataset and Azure 2019. Simulation results show a significant improvement in service-level agreement execution compared to comparative methods. Additionally, a reduction in energy consumption has been observed, while the average number of virtual machine migrations shows significant improvement.