9p virtual file system on qemu unstable

Bug #960923 reported by max on 2012-03-21
This bug affects 1 person
Affects Status Importance Assigned to Milestone
qemu-kvm (Ubuntu)

Bug Description

Hi Everyone,

The significant performance degradation in ubuntu 12.04 in work with 9p file system. It is working only for a couple of hours under intesive read/write operations. after this all operations stops. The system required complete shutdown of qemu and restarting it. the bug is qemu because the host is working well and other virtual machines connected to the same folder works without any problems.

In my case this problem is reproducible: 3 times in a raw.

guest:~$ uname -a
Linux biouml-1 3.2.0-19-virtual #30-Ubuntu SMP Fri Mar 16 18:07:13 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
host:~$ uname -a
Linux s0 3.2.0-19-generic #30-Ubuntu SMP Fri Mar 16 16:27:15 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux

kvm --version
QEMU emulator version 1.0 (qemu-kvm-1.0), Copyright (c) 2003-2008 Fabrice Bellard

VirtFS on /srv/shared type 9p (rw,trans=virtio,version=9p2000.L)

$ ps aux| grep kvm

 one-52 -uuid 50994459-9e97-b488-2893-b4a75a65c559 -nodefconfig -nodefaults -chardev socket,id=charmonitor,path=/var/lib/libvirt/qemu/one-52.monitor,server,nowait -mon chardev=charmonitor,id=monitor,mode=control -rtc base=utc -no-shutdown -drive file=/var/lib/one/var//52/images/disk.0,if=none,id=drive-virtio-disk0,format=raw,cache=writeback -device virtio-blk-pci,bus=pci.0,addr=0x5,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1 -drive file=/var/lib/one/var//52/images/disk.1,if=none,id=drive-virtio-disk4,format=raw,cache=writeback -device virtio-blk-pci,bus=pci.0,addr=0x6,drive=drive-virtio-disk4,id=virtio-disk4 -fsdev local,security_model=mapped,id=fsdev-fs0,path=/tank/biouml-shared -device virtio-9p-pci,id=fs0,fsdev=fsdev-fs0,mount_tag=VirtFS,bus=pci.0,addr=0x3 -netdev tap,fd=18,id=hostnet0 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=02:00:c0:a8:02:1e,bus=pci.0,addr=0x4 -usb -device usb-mouse,id=input0 -vnc -vga cirrus -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x7 -cpu host
ApportVersion: 2.0-0ubuntu2
Architecture: amd64
DistroRelease: Ubuntu 12.04
InstallationMedia: Ubuntu-Server 11.10 "Oneiric Ocelot" - Release amd64 (20111011)
MachineType: Supermicro H8QG6
NonfreeKernelModules: zfs zcommon znvpair zavl zunicode
Package: qemu-kvm 1.0+noroms-0ubuntu10 [origin: Ubuntu]
PackageArchitecture: amd64
ProcKernelCmdLine: BOOT_IMAGE=/vmlinuz-3.2.0-21-generic root=UUID=39e3c519-7ef2-44ea-af9f-60f5328337f2 ro
ProcVersionSignature: Ubuntu 3.2.0-21.34-generic 3.2.13
Tags: precise third-party-packages
Uname: Linux 3.2.0-21-generic x86_64
UnreportableReason: This is not an official Ubuntu package. Please remove any third party package and try again.
UpgradeStatus: Upgraded to precise on 2012-03-15 (18 days ago)
UserGroups: adm admin cdrom dialout kvm libvirtd lpadmin plugdev sambashare
dmi.bios.date: 10/28/2011
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 2.0a
dmi.board.asset.tag: 1234567890
dmi.board.name: H8QG6
dmi.board.vendor: Supermicro
dmi.board.version: 1234567890
dmi.chassis.asset.tag: 1234567890
dmi.chassis.type: 17
dmi.chassis.vendor: Supermicro
dmi.chassis.version: 1234567890
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr2.0a:bd10/28/2011:svnSupermicro:pnH8QG6:pvr1234567890:rvnSupermicro:rnH8QG6:rvr1234567890:cvnSupermicro:ct17:cvr1234567890:
dmi.product.name: H8QG6
dmi.product.version: 1234567890
dmi.sys.vendor: Supermicro

Serge Hallyn (serge-hallyn) wrote :

Thanks for reporting this bug.

It sounds like a memory leak. I don't see any commits looking like 9pfs memory leaks upstream. Do you know if you can reproduce this with qemu.git?

Can you explain exactly how to reproduce this (i.e. how to exercise 9p)?

Changed in qemu (Ubuntu):
importance: Undecided → Medium
status: New → Incomplete
max (mikhmv) wrote :

Hi Serge,

I have bad experience with compiling qemu from git. After this I was have to spent too much time to make it work with libvirt, seabios and so. I cannot shutdown our production machine for several days .... but I can try it if it will be pre-compiled.

I am using it for a data analysis of massive data. I think there are no way to transfer 200GB data and setup several programs to check it.

I can perform testes and collect data for it.

Serge Hallyn (serge-hallyn) wrote :

Is it possible for you to run

apport-collect 960923


I will try to reproduce.

Changed in qemu (Ubuntu):
status: Incomplete → New
affects: qemu (Ubuntu) → qemu-kvm (Ubuntu)
max (mikhmv) wrote :

I did it.

it asked me on launchpad login, authorization and time for access.

Did it sent you anything?
Can I provide more information?

max (mikhmv) wrote :

Additional confirmation that it is 9p Failure. When system failed after reboot VM without turning off qemu I receiving message on a picture.
It is working if I turn off qemu process and restart it again.

max (mikhmv) wrote : BootDmesg.txt

apport information

tags: added: apport-collected precise third-party-packages
description: updated
max (mikhmv) wrote : CurrentDmesg.txt

apport information

max (mikhmv) wrote : Dependencies.txt

apport information

max (mikhmv) wrote : KvmCmdLine.txt

apport information

max (mikhmv) wrote : Lspci.txt

apport information

max (mikhmv) wrote : Lsusb.txt

apport information

max (mikhmv) wrote : ProcCpuinfo.txt

apport information

max (mikhmv) wrote : ProcEnviron.txt

apport information

apport information

max (mikhmv) wrote : ProcModules.txt

apport information

apport information

max (mikhmv) wrote : UdevDb.txt

apport information

max (mikhmv) wrote : UdevLog.txt

apport information

Serge Hallyn (serge-hallyn) wrote :

(Awaiting info in bug 965867)

Changed in qemu-kvm (Ubuntu):
status: New → Incomplete
max (mikhmv) wrote :

VM is stable under heavy load.

Changed in qemu-kvm (Ubuntu):
status: Incomplete → Triaged
tags: added: fixed-upstream
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers