Enhancing Ethernet''s Reliability and Scalability

Ethernet is pervasive. This is due in part to its ease of use. Equipment can be added to an Ethernet network with little or no manual configuration. Furthermore, Ethernet is self-healing in the event of equipment failure or removal. Unfortunately, it suffers from significant reliability and scalability problems. In this book, we present different Ethernet reliability problems, characterizing their causes and effects. Then, we introduce two techniques to handle these problems. First technique is a fix to Ethernet's packet forwarding protocol - the Rapid Spanning Tree Protocol (RSTP). The second technique is a new device, called EtherFuse, we introduce that it can be inserted transparently into an existing Ethernet. In the event of a failure, EtherFuse detects it and mitigates its effects. Finally, we experimentally demonstrate the effectiveness of both techniques. Ethernet also faces scalability challenges due to broadcast traffic. To scale the Ethernet, we introduce another device, called EtherProxy, that uses caching to suppress broadcast traffic. EtherProxy is also backward compatible to existing Ethernet. Similarly, we experimentally demonstrate its effectiveness.

Autorentext

Khaled Elmeleegy is a researcher at Yahoo!Research. He earned his Ph.D. and M.S. in Computer Science from Rice University in 2009 and 2004 respectively. His work spans a number of different areas in computer systems, including networked systems, operating and distributed systems, and cloud computing.