From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: Why do programs freeze with big network transfers? Date: Thu, 30 Dec 2010 08:58:39 +0100 Message-ID: <1293695919.7150.18.camel@edumazet-laptop> References: Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: linux-kernel@vger.kernel.org, netdev To: Adam Nielsen Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Le jeudi 30 d=C3=A9cembre 2010 =C3=A0 17:25 +1000, Adam Nielsen a =C3=A9= crit : > Hi all, >=20 > I'm a bit stuck on this problem so I hope someone can help. My deskt= op PC is=20 > running kernel 2.6.33.1 and when I copy some largish files (2-3GB eac= h) onto=20 > an NFS share my PC becomes unusable, pretty much locking up for 60 se= conds at=20 > a time. >=20 > Everything works fine for a little while once the copy has begun - th= e files=20 > are read off the software-RAID-0 disks at about 200MB/sec, then after= 10=20 > seconds or so data starts going across the gigabit network at about 4= 0MB/sec=20 > (speed limited by the target system which pegs at 100% CPU due to lac= k of=20 > jumbo packets.) >=20 > After a few seconds of data going over the network, X-Windows freezes= =2E No=20 > screen updates, the mouse cursor won't move, for all intents and purp= oses the=20 > system has frozen solid. I'm playing music with XMMS2 and that keeps= going,=20 > but occasionally even that stops too. After a minute (between 45 and= 65=20 > seconds) everything unfreezes and keeps going as per normal. Less th= an 10=20 > seconds later everything freezes again for another minute! This keep= s going=20 > until the file transfer has finished. >=20 > When things unfreeze the disk is idle, and within 10 seconds the disk= starts=20 > up again and almost immediately the next minute-long freeze begins. = While=20 > things are frozen the network transfer continues, and bizarrely I can= log in=20 > to the machine over SSH where everything seems normal. 'top' reports= most=20 > processes are idle, and running a command line XMMS2 client happily r= eports=20 > that the song I am listening to is stuck at exactly the same point un= til the=20 > freeze is over, when the seconds start counting up again. >=20 > The reason I am stuck is that nothing is appearing in dmesg, so it ap= pears the=20 > kernel is unaware of the problem. Has anyone seen anything like this= before?=20 > I'm not sure what to do next. >=20 > Disks are connected to an Intel ICH9 SATA controller in AHCI mode, LA= N is a=20 > Realtek 8169, video card is nVidia GeForce 8600. Perhaps some combin= ation of=20 > this is to blame? >=20 > I have tried using cat to read these files into /dev/null and the sys= tem will=20 > happily read the files at full speed without freezing, and I have use= d ttcp's=20 > speed test function to send data over the network at full speed, whic= h also=20 > works without X11 freezing. Doing this at the same time (reading fro= m the=20 > disk and sending network traffic) also works fine without locking up,= so it=20 > seems the problems only arise when NFS gets involved. >=20 > 'mount' reports the options on the NFS share as: > rw,user=3Dadam,tcp,soft,intr,timeo=3D20,vers=3D3,addr=3D192.168.0.= 6 >=20 > Any suggestions about what I can do next? >=20 > Many thanks, > Adam. CC netdev This rings a bell here, could you try to apply commit=20 482964e56e1320cb7952faa1932d8ecf59c4bf75 (net: Fix the condition passed to sk_wait_event()) This commit was included in 2.6.36, so you could also try 2.6.36.2 kernel. http://git2.kernel.org/?p=3Dlinux/kernel/git/torvalds/linux-2.6.git;a=3D= commitdiff;h=3D482964e56e1320cb7952faa1932d8ecf59c4bf75 Thanks