Disentangling Flaws in Linux DCTCP

TL;DR

This paper systematically investigates performance issues in Linux DCTCP, identifying multiple bugs and flaws through experiments, and evaluates solutions to improve its reliability and performance.

Contribution

It provides a detailed analysis of the complex bugs affecting Linux DCTCP and proposes solutions, which is a novel systematic approach to debugging this protocol.

Findings

Multiple performance problems are caused by complex bugs.

Systematic experiments can identify root causes of network protocol flaws.

Proposed solutions improve DCTCP performance and reliability.

Abstract

In the process of testing improvements to the Linux DCTCP code in various scenarios, we found different performance problems kept surfacing with no apparent pattern. This report records a systematic sequence of experiments designed to track down the causes of these problems, which were found to be due to a complex tangle of bugs and flaws. The report also provides and evaluates solutions in each case.