Hybrid positioning method for data duplicate in data network system

Abstract

The invention belongs to mesh data management and sharing technical area. Characters are that self-similar layered isomorphic architecture is adopted in the system; granularity of node in the system can be adjusted arbitrarily. Thus, flexibility and extensibility of system are increased. Moreover, concept of super node is introduced in the system. Architecture full connection is adopted among these super nodes. The architecture reduces spending brought by huge route table to be maintained among nodes as well as influence on system caused by adding and quitting node frequently. The system includes two kinds of data: local metadata and global metadata. Global metadata makes storing and transferring data simpler; but user does not need to change custom of use. Parsing method in step-by-step is adopted for searching metadata. Thus, message for querying location of copy is limited in local area as far as possible.

Description

technical field [0001] The hybrid positioning method of data copy in the data grid system belongs to the field of distributed technology and system, especially the field of data management and sharing technology involved therein. Background technique [0002] The continuous development of network technology and distributed systems provides a more effective theoretical basis and technical means for distributed resource sharing of autonomous organizations in many application fields. Realizing resource sharing within an organization with the help of the Internet has become an effective way to solve resource sharing in the traditional sense, but it also brings new challenges. For example, different organizations and users distributed in various places need to share data through the network. These data have the following characteristics: First, they are heterogeneous, including data form, data size, and metadata. These factors make data sharing difficult. Second, these data ar...

AI Technical Summary

Problems solved by technology

The nodes in the peer-to-peer network are very dynamic, and nodes may join or exit the system at any time

[0008] Third, for unstructured P2P, nodes in the system can establish connection relationships with any other nodes. These connection relationships are determined by the system protocol and cannot be controlled by the system administrator. The system formed can meet the requirements, but for the data grid, which is a system composed of various subsystems, the connection and topological relationship in the system must be controlled.

[0009] Fourth, while emphasizing the role equality of participating nodes, the P2P system ignores the structural nature of the network organization

In this way, the communication of nodes in the same subnet may need to cross other subnets, resulting in increased communication delay and waste of bandwidth.

Images