TCP/IP protocol processing has become the dominant overhead in network servers. With the advent of memory-mapped SAN technologies such as VIA and InfiniBand, the intra-server transport protocols have been made extremely efficient. Our solution to reduce the client-server network processing overhead is to offload the TCP/IP protocol from the application hosts to dedicated processors, nodes or intelligent network interface. We called this architecture a TCP Server. In this paper we propose a Memory-Mapped Networking API and protocol (MemNet) to enable server applications to exploit the TCP Server architecture efficiently. The idea is to use low-overhead memory-mapped communication between application hosts and the TCP server to tunnel the so...
We propose a SAN architecture called Queue Pair IP (QPIP) that combines the interface from industry ...
User-level network interfaces allow applications direct access to the network without operating syst...
Efficient load balancing is essential for parallel distributed computing. Many parallel computing en...
TCP Server is a system architecture aiming to offload network processing from the host(s) running an...
TCP Server is a system architecture aiming to offload network processing from the host(s) running an...
This thesis will investigate the possibility offered by MEC to create novel ser- vices. In particula...
Ethernet continues to be the most widely used network architecture today due to its low cost and bac...
Masters ThesisNetworking servers, such as web servers, have been widely deployed in recent years. ...
Network applications often define policies to manage network traffic based on its attributes (e.g., ...
Networking servers, such as web servers, have been widely deployed in recent years. While developmen...
Server network performance is increasingly dominated by poorly scaling operations such as I/O bus cr...
Although many research efforts have been devoted to the network congestion against an increase of th...
Server load balancing technology has obtained much attention as much business proceeded towards e-co...
[[abstract]]This paper proposes a Web-based distributed network management architecture in which the...
The ServerNet-II System Area Network (SAN) is Compaq's native hardware implementation of the Vi...
We propose a SAN architecture called Queue Pair IP (QPIP) that combines the interface from industry ...
User-level network interfaces allow applications direct access to the network without operating syst...
Efficient load balancing is essential for parallel distributed computing. Many parallel computing en...
TCP Server is a system architecture aiming to offload network processing from the host(s) running an...
TCP Server is a system architecture aiming to offload network processing from the host(s) running an...
This thesis will investigate the possibility offered by MEC to create novel ser- vices. In particula...
Ethernet continues to be the most widely used network architecture today due to its low cost and bac...
Masters ThesisNetworking servers, such as web servers, have been widely deployed in recent years. ...
Network applications often define policies to manage network traffic based on its attributes (e.g., ...
Networking servers, such as web servers, have been widely deployed in recent years. While developmen...
Server network performance is increasingly dominated by poorly scaling operations such as I/O bus cr...
Although many research efforts have been devoted to the network congestion against an increase of th...
Server load balancing technology has obtained much attention as much business proceeded towards e-co...
[[abstract]]This paper proposes a Web-based distributed network management architecture in which the...
The ServerNet-II System Area Network (SAN) is Compaq's native hardware implementation of the Vi...
We propose a SAN architecture called Queue Pair IP (QPIP) that combines the interface from industry ...
User-level network interfaces allow applications direct access to the network without operating syst...
Efficient load balancing is essential for parallel distributed computing. Many parallel computing en...