Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161924AbbKTDPB (ORCPT ); Thu, 19 Nov 2015 22:15:01 -0500 Received: from mail-io0-f181.google.com ([209.85.223.181]:34341 "EHLO mail-io0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1161388AbbKTDO7 (ORCPT ); Thu, 19 Nov 2015 22:14:59 -0500 MIME-Version: 1.0 In-Reply-To: <1447762069.22599.88.camel@edumazet-glaptop2.roam.corp.google.com> References: <1447762069.22599.88.camel@edumazet-glaptop2.roam.corp.google.com> Date: Fri, 20 Nov 2015 11:14:58 +0800 Message-ID: Subject: Re: Hit regression with TCP_TW REUSE/RECYCLE From: Ethan Zhao To: Eric Dumazet Cc: netdev , LKML Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1695 Lines: 58 Eric, On Tue, Nov 17, 2015 at 8:07 PM, Eric Dumazet wrote: > On Tue, 2015-11-17 at 14:35 +0800, Ethan Zhao wrote: >> Tested the same case with 4.4-RC1, it was fixed in 4.4-RC1. >> But don't know which commit fixed it. >> >> # echo 1024 65535 > /proc/sys/net/ipv4/ip_local_port_range >> # cat /proc/sys/net/ipv4/ip_local_port_range >> 1024 65535 >> # echo 1 > /proc/sys/net/ipv4/tcp_tw_reuse >> # cat /proc/sys/net/ipv4/tcp_tw_reuse >> 1 >> # echo 1 > /proc/sys/net/ipv4/tcp_tw_recycle >> # cat /proc/sys/net/ipv4/tcp_tw_reuse >> 1 >> # ./accept -n 5 -r & >> [1] 11866 >> # port: 7954 >> ninst: 5 >> reuseport: 1 >> >> # ./connect -i 127.0.0.1 -n 5 -d 10 >> 78578.50 >> # uname -a >> Linux localhost.localdomain 4.4.0-rc1 #49 SMP Tue Nov 17 15:04:18 KST >> 2015 x86_64 x86_64 x86_64 GNU/Linux > > Maybe run a reverse bisection ? Hmmm, will do. > > tcp_tw_reuse and tcp_tw_recycle both set to one have never been > supported in linux. > > tcp_tw_recycle - BOOLEAN > Enable fast recycling TIME-WAIT sockets. Default value is 0. > It should not be changed without advice/request of technical > experts. > > tcp_tw_reuse - BOOLEAN > Allow to reuse TIME-WAIT sockets for new connections when it is > safe from protocol viewpoint. Default value is 0. > It should not be changed without advice/request of technical > experts. > > Thanks. > > Thanks, Ethan -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/