面向物理服务器集群的高性能自动化管控技术  

High-performance automated management and control technology for physical server clusters

在线阅读下载全文

作  者:张鹏 钱佳芮 王晓锋 王勇 奚相恺 邢嘉林 ZHANG Peng;QIAN Jia-rui;WANG Xiao-feng;WANG Yong;XI Xiang-kai;XING Jia-lin(Information Center,China Key System&Integrated Circuit Co.,Ltd.,Wuxi 214000,China;School of Artificial Intelligence and Computer Science,Jiangnan University,Wuxi 214122,China)

机构地区:[1]中科芯集成电路有限公司信息中心,江苏无锡214000 [2]江南大学人工智能与计算机学院,江苏无锡214122

出  处:《计算机工程与设计》2025年第4期1198-1207,共10页Computer Engineering and Design

基  金:国家自然科学基金项目(62172191)。

摘  要:为高效管理物理服务器集群并支持复杂计算任务,提出一种面向物理服务器集群的高性能自动化管控技术(HAMC)。针对服务器的异构性,设计一体化管理技术,实现硬件资源的高效统一管理;针对操作系统的差异性,采用基于异步消息的快速部署技术,实现操作系统的无人值守高效安装。优化传统网络架构,提高网络吞吐量。实验结果表明,HAMC可实现Inspur、H3C、Lenovo、DELL等品牌服务器的一体化管理,支持RedHat、CentOS、Ubuntu等系统的自动安装。与Cobbler和OpenStack相比,HAMC在操作系统部署效率上有所提升,在多跳和并发测试中,其网络吞吐量优于传统网络架构。To efficiently manage physical server clusters and support complex computational tasks,a high-performance automated management and control(HAMC)framework was proposed for physical server clusters.To address the heterogeneity of servers,an integrated management technique was designed to achieve efficient and unified hardware resource management.For the diversity of operating systems,an asynchronous message-based rapid deployment technique was adopted to enable unattended and efficient OS installation.The traditional network architecture was optimized to enhance the network throughput.Experimental results demonstrate that HAMC enables unified management of servers from Inspur,H3C,Lenovo,and DELL,and supports automated installation of RedHat,CentOS,and Ubuntu operating systems.Compared to Cobbler and OpenStack,HAMC significantly improves OS deployment efficiency.In multi-hop and concurrent testing scenarios,the network throughput of HAMC outperforms that of traditional network architectures.

关 键 词:高性能计算 物理服务器集群 资源管理 自动部署 异构性 网络优化 一体化管理 

分 类 号:TP393[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象