SCALAR: Scalable Architecture for Web Replication

FUNDING: National Science Foundation CCR-9902023

Graduate Students:

Mehmet Sayal and Bin Chen

Project Description:

The goal of the project was to design and implement a system for fast and reliable HTTP service. Our Web++ system achieves high reliability by dynamically replicating Web data among multiple Web servers. The delivery of the requested data item is guaranteed, for as long as at least one server which contains the desired data item is available. Upon detection of a server failure, the system will satisfy the client's request in a manner transparent to the client by using another server which is known to contain the desired data item. Web++ is very extensible, in the sense that its architecture allows for implementing additional performance optimizations, some of which were implemented as part of the SCALAR project. For publications, please refer to the Database Systems Group page.

Software:

Web++ is a free software and it can be distributed, in the hope that it will be found useful. One should feel free to re-distribute it and implement different optimization techniques on top of the existing architecture. Note that the Web++ software is distributed "as is" (without any warranty). This, as well as any modification of the Web++ software is in accordance with and subject to the GNU GENERAL PUBLIC LICENCE and the Web++. We provide two zip files:

• Applet-Servlet Version
which works on Web browsers in a "usual" manner - the user clicks on links of web page(s) and manually submits the requests.
• Stand-Alone Client and Servlet Version
which can be used for automated tests where stand-alone clients generate request(s) traffic automatically.

Go Back

This page updated on June 16, 2004 by GPT