【摘要】： 數(shù)碼技術的快速發(fā)展，使得越來越多的三維動畫作品出現(xiàn)在大型的電影、卡通、電視和廣告媒體中。在動畫制作過程中，對復雜的場景特效進行渲染，需要耗費大量的時間，渲染速度過慢成為動畫制作的一個瓶頸。集群渲染系統(tǒng)便是解決動畫制作瓶頸、提高工作效率的有效手段。集群渲染系統(tǒng)的廣泛應用，系統(tǒng)中軟硬件資源也在不斷地增加，因此系統(tǒng)失效的概率也逐漸增大。這樣，對系統(tǒng)進行監(jiān)控顯得尤為重要，監(jiān)控系統(tǒng)得到的數(shù)據(jù)可以用于負載均衡、任務調度和相應服務的調整，及時通知管理員有關硬件方面的故障或軟件的失效，以便保證集群的可靠性。 本文首先分析了集群渲染系統(tǒng)的構成、工作流程和特點，并針對集群渲染系統(tǒng)的工作流程指出了可視化技術在集群渲染監(jiān)控中的應用：系統(tǒng)狀態(tài)信息可視化監(jiān)控、作業(yè)狀態(tài)可視化監(jiān)控以及渲染結果可視化監(jiān)控。 其次，提出了集群渲染監(jiān)控的總體要求，介紹了用戶訪問系統(tǒng)的流程；指出本系統(tǒng)的主要功能要求，依據(jù)這些要求設計了系統(tǒng)的功能模塊，其中主要介紹了負載信息收集和作業(yè)信息收集模塊。 最后，具體介紹了監(jiān)測信息的獲取、存儲，，以及圖形化顯示，在LSF平臺上實現(xiàn)以圖形化方式實時顯示集群狀態(tài)、作業(yè)運行狀態(tài)、以及渲染結果等功能。
[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.
【學位授予單位】：華中師范大學
【學位級別】：碩士
【學位授予年份】：2007
【分類號】：TP277

【引證文獻】

相關碩士學位論文 前2條

1 王鈺;基于有機計算的動漫渲染集群系統(tǒng)管理技術的研究與應用[D];山東科技大學;2010年

2 琚攀攀;基于Web的高性能計算集群管理系統(tǒng)的設計與實現(xiàn)[D];西安建筑科技大學;2010年

本文編號：2375389