Overview
DYSWIS stands for Do You See What I See. This is a distributed automatic fault detection and diagnosis system. Traditional network diagnosis tools have been focused on server-side management. However, with such methods, it is hard to detect and diagnose network faults, which are caused by the particular environment of the end user.
Since it is running on the end user systems, DYSWIS can collect rich information from the end user system which detected the fault. Also, DYSWIS collaborates with other end users to see different views from other sides of network. It will ask other nodes to observe servers, routers or links. For example, they might request another node to try same transaction which caused the fault to see whether same fault occurs at the node.
Finally, DYSWIS analyze all the information obtained from the node itself as well as other nodes to figure out the original cause of the network problem.
Architecture
Features
- P2P-based (DHT)
- Remote probing (XML-RPC)
- Modularity (OSGi framework)
- Rule-based (Jess)
Implementation
Publications / Documents
Members
| Henning Schulzrinne |
Professor, Columbia University |
hgs@cs.columbia.edu |
| Kai Miao |
Researcher, Intel |
kai.miao@intel.com |
| Kyung Hwa Kim |
Ph.D. student, Columbia University |
kk2515@columbia.edu |
| Vishal Singh |
Ph.D. student, Columbia University |
vs2140@gmail.com |
Download