2010Unpublished venueRequires access

ICTCP: Incast Congestion Control for TCP

Haitao Wu, Zhenqian Feng, Chuanxiong Guo, Yongguang Zhang

Open publisher page 9 citations

Abstract

TCP incast congestion happens in high-bandwidth and lowlatency networks, when multiple synchronized servers send data to a same receiver in parallel [15]. For many important data center applications such as MapReduce[5] and Search, this many-to-one traffic pattern is common. Hence TCP incast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationship among TCP throughput, round trip time (RTT) and receive window. Different from the previous approach to mitigate the impact of incast congestion by a fine grained timeout value, our idea is to design an ICTCP (Incast congestion Control for TCP) scheme at the receiver side. In particular, our method adjusts TCP receive window proactively

About this research paper

What this paper is about

TCP incast congestion happens in high-bandwidth and lowlatency networks, when multiple synchronized servers send data to a same receiver in parallel [15]. For many important data center applications such as MapReduce[5] and Search, this many-to-one traffic pattern is common. Hence TCP incast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationship among TCP throughput, round trip time (RTT) and receive window. Different from the previous approach to mitigate the impact of incast congestion by a fine grained timeout value, our idea is to design an ICTCP (Incast congestion Control for TCP) scheme at the receiver side. In particular, our method adjusts TCP receive window proactively

Why it matters

OpenAlex reports 9 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

TCP incast congestion happens in high-bandwidth and lowlatency networks, when multiple synchronized servers send data to a same receiver in parallel [15]. For many important data center applications such as MapReduce[5] and Search, this many-to-one traffic pattern is common. Hence TCP incast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationship among TCP throughput, round trip time (RTT) and receive window. Different from the previous approach to mitigate the impact of incast congestion by a fine grained timeout value, our idea is to design an ICTCP (Incast congestion Control for TCP) scheme at the receiver side. In particular, our method adjusts TCP receive window proactively

Key concepts: TCP Friendly Rate Control, TCP global synchronization, Computer network, TCP Westwood plus, Computer science, CUBIC TCP, TCP acceleration, Zeta-TCP

Related papers

Back to paper searchBrowse research topicsOriginal source
ICTCP: Incast Congestion Control for TCP — Research Paper | ScholarLens