分布式计算站点统一部署和监控系统的研究与实现

Research and Implementation of Unified Deployment and Monitoring System for Distributed-Sites

近几年,高能物理合作的分布式计算站点数量越来越多,分布越来越广。物理软件的运行需要有稳定可靠的计算环境和统一的系统软件配置。本文通过对分布式站点统一部署和监控方案进行研究,从分布式架构设计、软件配置的文件同步和监控数据的采集进行了研究和设计,实现站点统一管理和运维。该方案的研究和实现有利于分布式计算站点的统一配置和集中运维,减轻管理和运维成本,保障物理作业在各个站点的可靠运行。

In recently years,there are more and more distributed computing sites joined into researching of high energy physics.Physics analysis software require a more stable and unified computing environment. The article researched distributed architecture and synchronizing sot^ware configuration and monitoring data acquisition,in order to realize unified management and operations.The solution can benefit unified operation and maintenance for distributed sites to ensure physics jobs running correctly.