Configuration Management and Infrastructure Monitoring Using CFEngine and Icinga for Real-time Heterogeneous Data Taking Environment

21st International Conference on Computing in High Energy and Nuclear Physics

物理学;计算机科学

Poat, M.D.^1 ; Lauret, J.^1 ; Betts, W.^1

BrookhavenNational Laboratory, P.O. Box 5000, Upton

NY

11973-5000, United States^1

关键词: Computing infrastructures;    Configuration management;    Configuration management tools;    Cyber infrastructures;    Infrastructure monitoring;    Long term monitoring;    Real time data collections;    Sustainable solution;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/664/5/052020/pdf
DOI  :  10.1088/1742-6596/664/5/052020

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

The STAR online computing environment is an intensive ever-growing system used for real-time data collection and analysis. Composed of heterogeneous and sometimes groups of custom-tuned machines, the computing infrastructure was previously managed by manual configurations and inconsistently monitored by a combination of tools. This situation led to configuration inconsistency and an overload of repetitive tasks along with lackluster communication between personnel and machines. Globally securing this heterogeneous cyberinfrastructure was tedious at best and an agile, policy-driven system ensuring consistency, was pursued. Three configuration management tools, Chef, Puppet, and CFEngine have been compared in reliability, versatility and performance along with a comparison of infrastructure monitoring tools Nagios and Icinga. STAR has selected the CFEngine configuration management tool and the Icinga infrastructure monitoring system leading to a versatile and sustainable solution. By leveraging these two tools STAR can now swiftly upgrade and modify the environment to its needs with ease as well as promptly react to cyber-security requests. By creating a sustainable long term monitoring solution, the detection of failures was reduced from days to minutes, allowing rapid actions before the issues become dire problems, potentially causing loss of precious experimental data or uptime.