netdev - Re: [PATCH net-next v2] tcp: Ensure DCTCP reacts to losses

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 4 Apr 2019 18:10:42 +0000
From:   Lawrence Brakmo <brakmo@...com>
To:     David Miller <davem@...emloft.net>,
        "olivier.tilmans@...ia-bell-labs.com" 
        <olivier.tilmans@...ia-bell-labs.com>
CC:     "koen.de_schepper@...ia-bell-labs.com" 
        <koen.de_schepper@...ia-bell-labs.com>,
        "research@...briscoe.net" <research@...briscoe.net>,
        "fw@...len.de" <fw@...len.de>,
        "borkmann@...earbox.net" <borkmann@...earbox.net>,
        "ycheng@...gle.com" <ycheng@...gle.com>,
        "ncardwell@...gle.com" <ncardwell@...gle.com>,
        "edumazet@...gle.com" <edumazet@...gle.com>,
        "agshew@...il.com" <agshew@...il.com>,
        "glenn.judd@...ganstanley.com" <glenn.judd@...ganstanley.com>,
        "kuznet@....inr.ac.ru" <kuznet@....inr.ac.ru>,
        "yoshfuji@...ux-ipv6.org" <yoshfuji@...ux-ipv6.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH net-next v2] tcp: Ensure DCTCP reacts to losses


On 4/4/19, 10:53 AM, "David Miller" <davem@...emloft.net> wrote:

    From: "Tilmans, Olivier (Nokia - BE/Antwerp)" <olivier.tilmans@...ia-bell-labs.com>
    Date: Thu, 4 Apr 2019 12:24:02 +0000
    
    > RFC8257 §3.5 explicitly states that "A DCTCP sender MUST react to
    > loss episodes in the same way as conventional TCP".
    ...
    > This patch replace the clamping of alpha by setting ssthresh to
    > half of cwnd for both fast retransmissions and RTOs, at most once
    > per RTT. Consequently, the dctcp_clamp_alpha_on_loss module parameter
    > has been removed.
    
    Applied to 'net' and queued up for -stable, thanks.
    
DCTCP is meant to be used in environments where the switches/routers do ECN marking, so it is not surprising that it performs badly when used in environments where it was not meant to be used. Has anyone measured the effect of this changed when DCTCP is used in environments where all the switches/routers do ECN marking? My concern is that we could end up hurting performance when DCTCP is used how it was meant to be used in order to protect incorrect uses of DCTCP.