Digital Organisms Access To Resources And Load Balancing

With the development of Peer-to-Peer, The P2P system not only makes use of client hosts that are distributed in the network ,but also makes the best use of the client resource at the edge of the Internet。The clients may conveniently enjoy the services that P2P system provided. Owing to the heterogeneous among the service nodes (such as computing capability, memory size, disk space, network bandwidth) result in imbalace of visiting resource. In order to enable the system to obtain better performance, system load balancing is needed.On the basis of the development of the distributed system and the characteristics of P2P application mode, 8010 research Developed a digital organisms system. at present, the digital organisms system has implemented Storage and management of resource with P2P system. the nodes that composed of the digital organisms system also have heterogeneous, and therefore faced with the following problem: different nodes may lead to Load imbalance because of different processing capability.When the client access the resources in the digital organisms system. The system first do the meta-data storage location with DHT technology, after getting the resource meta-data, according to meta-data information, find the storage node corresponding to the resources. Finally transmit the data between the users and nodes which provide services. The scheduling involved in this process can provide optimal service, this is implemented by load balancing service scheduling module. At the same time, because the join and leave of the nodes is random, effectively and timely dealing with resource imbalance caused by the join and leave of the nodes is the main problem that Load balancing need to solve. We have designed Node heartbeat information monitoring modules and services register logout modules to solve the problem. Because different node have different computing capability, when load overloading appear, it is important to Take measures to balance the nodes load. So that the node in the system can provide services fairly as far as possible. In order to judge whether a node is overload, we designed alarm information modules. When the node is overload, alarm information modules are trigged and notify the system that load overload with a alarm message. When the system accepted the alarm message, it operated the load balancing that based on the minimum subset of k designed in the article. When clients access to the resources, in order to find the nearest node to provide service and avoid the other node lead to imbalance in the routing process because of their long distance, so as far as possible to find the nearest node to provide services. Based on this purpose, we designed routing distance judge module. The load balancing scheme is complement and further development for the resource management function in digital organism system, and is part of digital organism system resource management system. Finally we tested it, as the result of improving the storage of resources and the access of load balancing.