| Project/Area Number |
16500039
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Computer system/Network
|
|---|
| Research Institution | Osaka City University |
|---|
Principal Investigator |
HAYATO Ishibashi Osaka City University, Graduate School for Creative Cities, Associate Professor, 大学院創造都市研究科, 助教授 (70212925)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MATSUURA Toshio Osaka City University, Graduate School for Creative Cities, Professor, 大学院創造都市研究科, 教授 (40127296)
ABE Kota Osaka City University, Graduate School for Creative Cities, Associate Professor, 大学院創造都市研究科, 助教授 (40291603)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥3,500,000 (Direct Cost: ¥3,500,000)
Fiscal Year 2006: ¥1,000,000 (Direct Cost: ¥1,000,000)
Fiscal Year 2005: ¥1,300,000 (Direct Cost: ¥1,300,000)
Fiscal Year 2004: ¥1,200,000 (Direct Cost: ¥1,200,000)
|
|---|
| Keywords | Distributed filesystem / Virtual file server / Distributed backup system / Distributed hash table / Hierarchical clustering / Peer-to-peer / Distributed autonomous system / 階層的クラスタリング / 仮想ファイルサーバ / ネットワークファイルシステム / 完全分散型自律分散システム / クラスタリング / 自律分散型 |
|---|
| Research Abstract |
This project targeted to develop a virtual fileserver system, which consists of many client PCs as its component nodes. Each client PC provides a part of its own storage area to construct a virtual filesystem that supplies a large amount of shared disk space consequently. One of the most important benefits of this approach is as the number of clients increases, the total capacity and total processing power increase as well. All nodes in the system run by a distributed, autonomous manner so that the entire system does not stop working if some portion of nodes got failure. File segments are redundantly stored for robustness.
To archive this project, we set a design of a distributed file backup system as a first goal. It might be easier to implement than filesystem because backup systems are not so sensitive from the aspect of performance requirements. Then we have designed and implemented a distributed virtual filesystem with the knowledge acquired from the first goal.
We also studied on distributed hash tables (DHT). Since each node in the system works autonomously and may join/leave at anytime, flexibility of DHTs is favorable to manage nodes and file placement. Our proposed DHT performs with our hierarchical clustering algorithm. It makes the utilization of network resource as effectively as possible by traffic localization.
|
