It, the collider of the high energy physics area, produces the huge amount of data. The data then should be dispersed over the world in high speed and in high capacity, for the purpose of analyzing the data by the scientists in the world. So they built and organized Tier system, a global scale infrastructure for transferring the data in hierarchical manners over the world. The scientists have to connect to the each final Tier system in their each country, and have to download and analyze the part of the data for their HEP researches. So, they need a storage system to keep the large amount of the data, and also need to run tools to analyze the data.
In this poster, we suggest a system that can provide a number of nodes for the use for saving and analyzing data. The system is a kind of cluster machine where the nodes can be shared by any number of users unless they need them simultaneously. The users can get assigned the nodes as many as they request, and they can keep the data in the nodes as long as they want to. The cluster system has programmable switches which can connect the nodes in any topology. This makes it possible to organize the nodes in star topology, in bus, in mesh, and so on. The users can request (1) the number of nodes, (2) the topology between them, (3) link bandwidth. The users can get full permission to use the nodes and this can give a great way to run their tools for analyzing. The data and any files, and even the last status of the environment they used, are managed, so they can continue their job very easily for the next visit.
This solution can perform the jobs which should be done at the scientists' terminal system for HEP researches. The system can give any number of the node by the request through web, and can give full permission of the node. The scientist can work the HEP researches well and efficiently with this solution. The scientists can do new challenges which were not tried due to the limitation of the system environment.