RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
Bug 1374623 - RHSA-2016-1756 breaks migration of instances
Summary: RHSA-2016-1756 breaks migration of instances
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat Enterprise Linux 7
Classification: Red Hat
Component: qemu-kvm-rhev
Version: 7.3
Hardware: Unspecified
OS: Unspecified
urgent
high
Target Milestone: rc
: ---
Assignee: Stefan Hajnoczi
QA Contact: huiqingding
URL:
Whiteboard:
Depends On: 1372763 1376542
Blocks:
TreeView+ depends on / blocked
 
Reported: 2016-09-09 08:50 UTC by Marcel Kolaja
Modified: 2021-08-30 12:21 UTC (History)
35 users (show)

Fixed In Version: qemu-kvm-rhev-2.3.0-31.el7_2.22
Doc Type: Bug Fix
Doc Text:
The fix for CVE-2016-5403 caused migrating guest instances to fail with a "Virtqueue size exceeded" error message. With this update, the value of the virtualization queue is recalculated after the migration, and the described problem no longer occurs.
Clone Of: 1372763
Environment:
Last Closed: 2016-11-17 15:01:23 UTC
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Red Hat Knowledge Base (Solution) 2598111 0 None None None 2016-09-09 08:50:30 UTC
Red Hat Product Errata RHBA-2016:2803 0 normal SHIPPED_LIVE qemu-kvm-rhev bug fix and enhancement update 2016-11-17 20:01:03 UTC

Description Marcel Kolaja 2016-09-09 08:50:04 UTC 
This bug has been copied from bug #1372763 and has been proposed
to be backported to 7.2 z-stream (EUS).
Comment 4 Miroslav Rezanina 2016-09-13 12:49:44 UTC 
Fix included in qemu-kvm-rhev-2.6.0-25.el7
Comment 5 Miroslav Rezanina 2016-09-14 08:08:16 UTC 
Fix included in qemu-kvm-rhev-2.3.0-31.el7_2.22
Comment 7 lijin 2016-09-22 05:43:13 UTC 
with qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64,I still hit "Virtqueue size exceeded" error when do migration during disk(blk) io stress.

steps:
1.boot win8-32 guest with virtio-blk-pci device:
-object iothread,id=thread0 -drive file=win8-32-rhel7u2.raw,if=none,id=drive-ide0-0-0,format=raw,serial=mike_cao,cache=none -device virtio-blk-pci,iothread=thread0,drive=drive-ide0-0-0,id=ide0-0-0,bootindex=1 

2.run CrystalDiskMark in guest

3.do migration

actual result:
migration failed with:
src:(qemu) qemu-kvm: Virtqueue size exceeded
dst:(qemu) qemu-kvm: error while loading state section id 2(ram)
qemu-kvm: load of migration failed: Input/output error
Comment 8 Stefan Hajnoczi 2016-09-22 13:47:58 UTC 
(In reply to lijin from comment #7)
> with qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64,I still hit "Virtqueue size
> exceeded" error when do migration during disk(blk) io stress.
> 
> steps:
> 1.boot win8-32 guest with virtio-blk-pci device:
> -object iothread,id=thread0 -drive
> file=win8-32-rhel7u2.raw,if=none,id=drive-ide0-0-0,format=raw,
> serial=mike_cao,cache=none -device
> virtio-blk-pci,iothread=thread0,drive=drive-ide0-0-0,id=ide0-0-0,bootindex=1 
> 
> 2.run CrystalDiskMark in guest
> 
> 3.do migration
> 
> actual result:
> migration failed with:
> src:(qemu) qemu-kvm: Virtqueue size exceeded
> dst:(qemu) qemu-kvm: error while loading state section id 2(ram)
> qemu-kvm: load of migration failed: Input/output error

I was able to reproduce this problem with a Linux guest running fio.  It is a different bug since the error happens on the source QEMU while the patch for this BZ fixes the destination QEMU.
Comment 9 lijin 2016-09-23 07:53:09 UTC 
(In reply to Stefan Hajnoczi from comment #8)
> I was able to reproduce this problem with a Linux guest running fio.  It is
> a different bug since the error happens on the source QEMU while the patch
> for this BZ fixes the destination QEMU.

Thanks,I will report a new bug to track it.
Comment 10 lijin 2016-09-23 08:05:39 UTC 
This issue is reproduced on rhel7.2-z(qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64),can NOT reproduce with rhel7.3 latest version(qemu-kvm-rhev-2.6.0-26.el7.x86_64)

As Z stream bug should clone from Y stream bug,I'm a little confused how to handle it?
Comment 11 Stefan Hajnoczi 2016-09-23 08:33:01 UTC 
(In reply to lijin from comment #10)
> This issue is reproduced on
> rhel7.2-z(qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64),can NOT reproduce with
> rhel7.3 latest version(qemu-kvm-rhev-2.6.0-26.el7.x86_64)
> 
> As Z stream bug should clone from Y stream bug,I'm a little confused how to
> handle it?

Right, code inspection shows that RHEL 7.3 and upstream QEMU do not suffer from this race condition.  So we need a 7.2.z-only BZ.

I don't know the process either but I've asked on IRC.  Will update the BZ when I receive an answer.
Comment 12 Stefan Hajnoczi 2016-09-23 09:15:29 UTC 
(In reply to Stefan Hajnoczi from comment #11)
> (In reply to lijin from comment #10)
> > This issue is reproduced on
> > rhel7.2-z(qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64),can NOT reproduce with
> > rhel7.3 latest version(qemu-kvm-rhev-2.6.0-26.el7.x86_64)
> > 
> > As Z stream bug should clone from Y stream bug,I'm a little confused how to
> > handle it?
> 
> Right, code inspection shows that RHEL 7.3 and upstream QEMU do not suffer
> from this race condition.  So we need a 7.2.z-only BZ.
> 
> I don't know the process either but I've asked on IRC.  Will update the BZ
> when I receive an answer.

09:46 < mrezanin> stefanha: As usual...create normal bz (for both z-stream and y-stream). Then we negotiate clone and after it y-stream is closed with proper marking and z-stream is solved
09:48 < stefanha> mrezanin: Does this mean: create a BZ with both y-stream and z-stream flags, then leave a comment saying it's relevant for z-stream only, close it as NOTABUG?
09:48 < mrezanin> stefanha: Yes...but close after cloning

I will create the BZ and CC you.
Comment 13 huiqingding 2016-10-10 03:06:34 UTC 
Reproduce this bug using the following version:
kernel-3.10.0-327.37.1.el7.x86_64
qemu-kvm-rhev-2.3.0-31.el7_2.21.x86_64

Reproduce steps:
1. create a 4M lv
# pvcreate /dev/sdg
# vgcreate testvg /dev/sdg
# lvcreate -L 4M -T testvg/testlv
# lvs
  LV     VG                  Attr       LSize   Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
  home   rhel_hp-dl380pg8-09 -wi-ao---- 212.61g                                                    
  root   rhel_hp-dl380pg8-09 -wi-ao----  50.00g                                                    
  swap   rhel_hp-dl380pg8-09 -wi-ao----  15.75g                                                    
  testlv testvg              twi-a-tz--   4.00m             0.00   0.88  

2. create a data disk image based on the above lv
# qemu-img create -f qcow2 /dev/testvg/testlv 10G

3. boot a rhel7.3 guest with the above data disk image
# /usr/libexec/qemu-kvm \
 -S \
 -name 'rhel7.3' \
 -machine pc \
 -m 4096 \
 -smp 4,maxcpus=4,sockets=1,cores=4,threads=1 \
 -cpu SandyBridge \
 -rtc base=localtime,clock=host,driftfix=slew \
 -nodefaults \
 -boot menu=on \
 -enable-kvm \
 -monitor stdio \
 -drive file=/mnt/rhel7.3.raw,format=raw,id=drive_sysdisk,if=none,cache=none,aio=native,werror=stop,rerror=stop \
 -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x8,drive=drive_sysdisk,bootindex=1 \
  -drive if=none,cache=none,format=qcow2,file=/dev/testvg/testlv,werror=stop,id=drive-virtio-disk0 \
 -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x9,drive=drive-virtio-disk0,id=virtio-disk0 \
 -vga qxl \
 -spice port=5900,disable-ticketing

4. on the same host, use the same command line with "-incoming tcp:0:5800", boot the rhel7.3 guest

5. inside guest
# dd if=/dev/zero of=/dev/vdb oflag=direct bs=4k

6. after guest is paused with io-error, do migration
(qemu) info status
VM status: paused (io-error)
(qemu) migrate -d tcp:0:5800

7. on host, grow the logical volume by 4 MB
# lvresize -L +4M /dev/testvg/testlv

8. in destination, resume the guest
(qemu)c

after step8, "Virtqueue size exceeded" error from destination QEMU and qemu-kvm quits.

Verify this bug using the following version:
kernel-3.10.0-327.37.1.el7.x86_64
qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64

Do the above test, after step 8, destination qemu-kvm did not quit and guest can resume normally.
Comment 14 huiqingding 2016-10-10 03:07:57 UTC 
Based on comment #13, set this bug to be verified.
Comment 15 Irina Petrova 2016-10-27 09:48:49 UTC 
Hey guys, 

Since qemu-kvm-rhev-2.3.0-31.el7_2.22.x86_64 obviously passed QA (c#13, c#14) when can we expect it?

I have another customer asking for it.

Regards,
Irina
Comment 23 errata-xmlrpc 2016-11-17 15:01:23 UTC 
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://rhn.redhat.com/errata/RHBA-2016-2803.html
Note You need to log in before you can comment on or make changes to this bug.