基于LSF平台的集群渲染可视化监控研究与实现
发布时间：2018-12-12 22:59
【摘要】： 数码技术的快速发展，使得越来越多的三维动画作品出现在大型的电影、卡通、电视和广告媒体中。在动画制作过程中，对复杂的场景特效进行渲染，需要耗费大量的时间，渲染速度过慢成为动画制作的一个瓶颈。集群渲染系统便是解决动画制作瓶颈、提高工作效率的有效手段。集群渲染系统的广泛应用，系统中软硬件资源也在不断地增加，因此系统失效的概率也逐渐增大。这样，对系统进行监控显得尤为重要，监控系统得到的数据可以用于负载均衡、任务调度和相应服务的调整，及时通知管理员有关硬件方面的故障或软件的失效，以便保证集群的可靠性。 本文首先分析了集群渲染系统的构成、工作流程和特点，并针对集群渲染系统的工作流程指出了可视化技术在集群渲染监控中的应用：系统状态信息可视化监控、作业状态可视化监控以及渲染结果可视化监控。 其次，提出了集群渲染监控的总体要求，介绍了用户访问系统的流程；指出本系统的主要功能要求，依据这些要求设计了系统的功能模块，其中主要介绍了负载信息收集和作业信息收集模块。 最后，具体介绍了监测信息的获取、存储，，以及图形化显示，在LSF平台上实现以图形化方式实时显示集群状态、作业运行状态、以及渲染结果等功能。
[Abstract]:With the rapid development of digital technology, more and more 3D animation works appear in large film, cartoon, TV and advertising media. In the process of animation making, it takes a lot of time to render the special effects of complex scenes, so the slow rendering speed becomes a bottleneck of animation production. Cluster rendering system is an effective way to solve the bottleneck of animation production and improve working efficiency. With the extensive application of cluster rendering system, the software and hardware resources in the system are increasing, so the probability of system failure increases gradually. In this way, it is particularly important to monitor the system. The data obtained from the monitoring system can be used for load balancing, task scheduling and adjustment of corresponding services, and timely notification to administrators of hardware failures or software failures. In order to ensure the reliability of the cluster. Firstly, this paper analyzes the composition, workflow and characteristics of cluster rendering system, and points out the application of visualization technology in cluster rendering monitoring: visual monitoring of system state information. Visual monitoring of job status and visual monitoring of rendering results. Secondly, the overall requirements of cluster rendering monitoring are put forward, and the flow of user access system is introduced. The main functional requirements of the system are pointed out. According to these requirements, the functional modules of the system are designed, in which the load information collection and the job information collection module are mainly introduced. Finally, the paper introduces the acquisition, storage and graphical display of monitoring information, and realizes the functions of real-time display of cluster state, job running state and rendering result on LSF platform.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2007
【分类号】：TP277
本文编号：2375389
[Abstract]:With the rapid development of digital technology, more and more 3D animation works appear in large film, cartoon, TV and advertising media. In the process of animation making, it takes a lot of time to render the special effects of complex scenes, so the slow rendering speed becomes a bottleneck of animation production. Cluster rendering system is an effective way to solve the bottleneck of animation production and improve working efficiency. With the extensive application of cluster rendering system, the software and hardware resources in the system are increasing, so the probability of system failure increases gradually. In this way, it is particularly important to monitor the system. The data obtained from the monitoring system can be used for load balancing, task scheduling and adjustment of corresponding services, and timely notification to administrators of hardware failures or software failures. In order to ensure the reliability of the cluster. Firstly, this paper analyzes the composition, workflow and characteristics of cluster rendering system, and points out the application of visualization technology in cluster rendering monitoring: visual monitoring of system state information. Visual monitoring of job status and visual monitoring of rendering results. Secondly, the overall requirements of cluster rendering monitoring are put forward, and the flow of user access system is introduced. The main functional requirements of the system are pointed out. According to these requirements, the functional modules of the system are designed, in which the load information collection and the job information collection module are mainly introduced. Finally, the paper introduces the acquisition, storage and graphical display of monitoring information, and realizes the functions of real-time display of cluster state, job running state and rendering result on LSF platform.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2007
【分类号】：TP277
【引证文献】
相关硕士学位论文 前2条
1 王钰;基于有机计算的动漫渲染集群系统管理技术的研究与应用[D];山东科技大学;2010年
2 琚攀攀;基于Web的高性能计算集群管理系统的设计与实现[D];西安建筑科技大学;2010年
本文编号：2375389
本文链接：https://www.wllwen.com/wenyilunwen/guanggaoshejilunwen/2375389.html
