Improving server application performance via pure TCP ACK receive optimization

Network stack performance is critical to server scalability and user-perceived application experience. Perpacket overhead is a major bottleneck in scaling network I/O. While much effort is expended on reducing perpacket overhead for data-carrying packets, small control packets such as pure TCP ACKs have received relatively scarce attention. In this paper, we show that ACK receive processing can consume up to 20% cycles in server applications. We propose a simple kernel-level optimization which reduces this overhead through fewer memory allocations and a simplified code path. Using this technique, we demonstrate cycles savings of 15% in a Web application, and 33% throughput improvement in reliable multicast.