From owner-freebsd-stable@freebsd.org Sun Sep 18 20:34:21 2016 Return-Path: Delivered-To: freebsd-stable@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 8C174BDFB84 for ; Sun, 18 Sep 2016 20:34:21 +0000 (UTC) (envelope-from hps@selasky.org) Received: from mail.turbocat.net (heidi.turbocat.net [88.198.202.214]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 529FD16C4 for ; Sun, 18 Sep 2016 20:34:20 +0000 (UTC) (envelope-from hps@selasky.org) Received: from laptop015.home.selasky.org (unknown [62.141.129.119]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by mail.turbocat.net (Postfix) with ESMTPSA id 1564E1FE023; Sun, 18 Sep 2016 22:34:12 +0200 (CEST) Subject: Re: 11.0 stuck on high network load To: Slawa Olhovchenkov References: <2490d030-e947-4842-8e91-6498864b0100@selasky.org> <20160918181016.GK2840@zxy.spb.ru> Cc: freebsd-stable@freebsd.org From: Hans Petter Selasky Message-ID: Date: Sun, 18 Sep 2016 22:38:58 +0200 User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:45.0) Gecko/20100101 Thunderbird/45.0 MIME-Version: 1.0 In-Reply-To: <20160918181016.GK2840@zxy.spb.ru> Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 18 Sep 2016 20:34:21 -0000 On 09/18/16 20:10, Slawa Olhovchenkov wrote: > On Sun, Sep 18, 2016 at 07:50:08PM +0200, Hans Petter Selasky wrote: > >> Hi, >> >> Got some tips regarding this thread. >> >> Some things you can try: >> >> 1) Compile kernel from projects/hps_head instead of your 11-stable? > > How many difference from 11-stable? Hi, The callout subsystem has a different implementation. Else identical. > >> 2) Set net.inet.tcp.per_cpu_timers=1 > > Already. From 10.x, by manual MFC. OK. > >> If the system just hangs, it is pretty likely that the timers are going >> in a loop due to typical use after free. >> >> Please keep me CC'ed, hence I'm not subscribed to @stable. >> --HPS