DeTail: Reducing the Flow Completion Time Tail in Datacenter Networks

David Zats, Tathagata Das, Prashanth Mohan, Dhruba Borthakur and Randy H. Katz

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2012-33
March 15, 2012

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2012/EECS-2012-33.pdf

Web applications have now become so sophisticated that rendering a typical page may require hundreds of intra-datacenter flows. At the same time, web sites must meet strict page creation deadlines of 200-300ms to satisfy user demands for interactivity. Long-tailed flow completion times make it challenging for web sites to meet these constraints. They are forced to choose between rendering a subset of the complex page, or delay its rendering, thus missing deadlines and sacrificing either quality or responsiveness. Either option leads to potential financial loss.

In this paper, we present a new cross-layer network stack aimed at reducing the long tail of flow completion times. The approach exploits cross-layer information to reduce packet drops, prioritize latency-sensitive flows, and evenly distribute network load, effectively reducing the long tail of flow completion times. We evaluate our approach through NS-3 based simulation and Click-based implementation demonstrating our ability to consistently reduce the tail across a wide range of workloads. We commonly achieve reductions of over 50\% in 99.9th percentile flow completion times without significantly impacting the median.


BibTeX citation:

@techreport{Zats:EECS-2012-33,
    Author = {Zats, David and Das, Tathagata and Mohan, Prashanth and Borthakur, Dhruba and Katz, Randy H.},
    Title = {DeTail: Reducing the Flow Completion Time Tail in Datacenter Networks},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {2012},
    Month = {Mar},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2012/EECS-2012-33.html},
    Number = {UCB/EECS-2012-33},
    Abstract = {Web applications have now become so sophisticated that rendering a typical page may require hundreds of intra-datacenter flows. At the same time, web sites must meet strict page creation deadlines of 200-300ms to satisfy user demands for interactivity. Long-tailed flow completion times make it challenging for web sites to meet these constraints. They are forced to choose between rendering a subset of the complex page, or delay its rendering, thus missing deadlines and sacrificing either quality or responsiveness. Either option leads to potential financial loss.

In this paper, we present a new cross-layer network stack aimed at reducing the long tail of flow completion times. The approach exploits cross-layer information to reduce packet drops, prioritize latency-sensitive flows, and evenly distribute network load, effectively reducing the long tail of flow completion times. We evaluate our approach through NS-3 based simulation and Click-based implementation demonstrating our ability to consistently reduce the tail across a wide range of workloads. We commonly achieve reductions of over 50\% in  99.9th percentile flow completion times without significantly impacting the median.}
}

EndNote citation:

%0 Report
%A Zats, David
%A Das, Tathagata
%A Mohan, Prashanth
%A Borthakur, Dhruba
%A Katz, Randy H.
%T DeTail: Reducing the Flow Completion Time Tail in Datacenter Networks
%I EECS Department, University of California, Berkeley
%D 2012
%8 March 15
%@ UCB/EECS-2012-33
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2012/EECS-2012-33.html
%F Zats:EECS-2012-33