1997 Fiscal Year Final Research Report Summary
Construction of Distributed Network Management System for High speed and Large Scale Information Network
| Project/Area Number |
08558033
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 展開研究 |
|---|
| Research Field |
情報システム学(含情報図書館学)
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
NEMOTO Yoshiaki G.S.I.S.Tohoku University Prof., 大学院・情報科学研究科, 教授 (60005527)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KATO Nei G.S.I.S.Tohoku University Assoc.Prof., 大学院・情報科学研究科, 助教授 (00236168)
MANSFIELD Glenn Institute of Sendai Applied Information Researcher, 主任研究員
HAMAMOTO Naokazu Communication Research Laboratory Researcher, 宇宙通信部, 室長
MAKINO Shozo Computer Center, Tohoku University Prof., 大型計算機センター, 教授 (00089806)
|
|---|
| Project Period (FY) |
1996 – 1997
|
| Keywords | Network-manugement / Distributed-management / SNMP / RMON |
|---|
| Research Abstract |
Computer networks expand in a very fast pace in recent year with diversity. How to manage the network for users becomes and important problem. In this research, we focus on the network fault management system. The target of this research is to search for the proper method of management information collection and develop a basic system of network fault management.
By the two years research and development, we proposed a new method of information collection for management system. By controlling management traffic in considering of the bandwidth and congestion of network this method can be used on any networks with different speed. We also proposed the idea of synchronization called Snapshot for obtaining high quality network management information. By using this idea, not only the synchronized management can be obtained easily on a SNMP (Simple Network Management Protocol) platform, but also the processing time is reduced.
By making use of this two idea, we developed a prototype of fault management system. In this system, we proposed divide and conquer algorithm for detecting faults occcurred in network. This algorithm used event-driven, dynamic symptom isolation method. We have evaluated the performance of this proposed algorithm and have examined the effect of the various operational parameters on two different kinds of networks. In comparison with the conventional system, the system with divide and conquer mechanism improved the performance significantly and at the same time reduced management cost. We consider that the purpose of this research have been achieved.
|
Research Products
(12 results)
