Research On Key Technologies Of Mass Storage Resource Management

With the development of science and technology, computing centralized information system is gradually transferring to data centralized information system. The name of information center transfers from“Computing Center”,“Network Center”to“Data Center”. This naming change lively indicates the change of the relative importance of the three key elements: processing, transfer and storage. However, comparing with computing and transferring, storage is always the bottleneck of the computing system. And moreover, the management method of mass storage resource is the key point of storage system. This research concentrates on mass storage resource management method under the network environment from three aspects. Firstly, an efficient storage management method through virtualization strategy was proposed to solve problems in isomerous mass storage resource management like huge amounts of storage nodes, diversity of device specifications and network components. An out-of-band virtualization scheme and prototype system in block-level was designed and realized. Secondly, to optimize the performance of the prototype system, a scheduling optimization algorithm was proposed for multi-user requests based on task queue merging. The simulation result shows that this algorithm can significantly increase the I/O request response ability of storage system.Thirdly, the research of mass storage resource management was expanded to wide area network (WAN) environment. Considering the new characteristics of data in WAN, a WAN storage resource management framework was proposed which integrated storage management, metadata management and file service. And we also established a Stochastic Petri Nets (SPN) model to analyze the performance.Base on previous works, this research obtains following achievements through theoretical research and practical development.According to the diversity of storage resources and complexity of storage networks of isomerous mass storage system, a standardization method was proposed to describe the isomerous mass storage resources. This standardization method supports the isomerous mass storage resources discovery, resource information maintenance and resource status monitoring. All of these formed in the base of storage virtualization. We designed and implemented an out-of-band virtualization scheme in block level; realized the transparent unified storage pool model for application and management; developed a storage access agent in file level which is transparent for the application; presented an aggregation and redirection technology for namespace. Then the unified management and access of isomerous mass storage resources based on the above methods was realized. In view of the highly complex challenges of the storage resources management mechanism in actual application environment, a highly efficient management method through virtualization strategy was proposed. This method realizes the dynamic and on-demand allocation for storage resources through open virtualization strategy language and related software integration.I/O request scheduling algorithm is an important factor which affects the performance of network storage system. According to the multi-user I/O request feature in storage system, an I/O request scheduling optimization algorithm (SOA) based on task queue and merging was proposed. The I/O request tasks of multiple network users were queued, and the data requested was sorted in certain order on logical disk. Then the continuous users’requests were merged and the arm of disk could move according to specific rule. This reduces the seek time and rotation delay significantly, and thus decreases the average response time and improves the network storage system’s I/O performance. On the other hand, the SOA algorithm combines the“task merge”and“cut in queue”mechanism. Base on this combination, the algorithm takes into consideration both the performance of the whole system and the response time of a single request. The algorithm also eliminates the phenomenon of“hungry”.The next generation Internet represented by IPv6 has increased the number of network users greatly. This demands not only various data, but also better usability, reliability, maintainability, and security, etc. Besides, it requires effective utilization of information in production, release, sharing and spread. According to these characteristics of the next generation Internet, the management methods of storage system and metadata, efficient use of storage space, and rapid transmission of data was investigated. And a WAN storage resourses management framework was built for next generation Internet. This framework used multi-layer and scalable distributed storage model. Connecting isomerous storage systems which act as basic storage nodes to next generation Internet through network interface, the framework builds a direct data transmission between storage nodes and users. It expands the capacity and speeds up the data transmission at the same time. This ensures efficient control and data management between servers and storage nodes. Its storage management follows SMI-S management standard. So it can provide standardized communication mode in WAN storage system. The framework’s intelligent metadata management methods ensure the quality of service and storage space utilization. Innovative design of the WAN storage middleware makes the system process the tasks of data transmission in complex network environment efficiently and allows for both high speed and security for WAN data transmission.