Bug 703112 - Ping-Pong live migration with network load cause guest network down
Summary: Ping-Pong live migration with network load cause guest network down
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat Enterprise Linux 5
Classification: Red Hat
Component: kvm
Version: 5.7
Hardware: x86_64
OS: Linux
high
medium
Target Milestone: rc
: ---
Assignee: Orit Wasserman
QA Contact: Virtualization Bugs
URL:
Whiteboard:
Depends On:
Blocks: Rhel5KvmTier2
TreeView+ depends on / blocked
 
Reported: 2011-05-09 10:32 UTC by Mike Cao
Modified: 2014-03-04 00:24 UTC (History)
11 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2012-04-10 10:31:06 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Description Mike Cao 2011-05-09 10:32:23 UTC
Description of problem:


Version-Release number of selected component (if applicable):
Host info:
# uname -r
2.6.18-260.el5
# rpm -q kvm
kvm-83-232.el5

Guest info:
kernel-2.6.32-131.0.13.el6.i686

How reproducible:
sometimes

Steps to Reproduce:
1.Start RHEL6.1 guest on Host t1
eg:/usr/libexec/qemu-kvm -M rhel5.6.0 -m 4G -smp 4,sockets=4,cores=1,threads=1 -name VM -uuid 02b8079c-cca0-ae65-c63a-b2aa4119497d -monitor stdio -no-kvm-pit-reinjection -usbdevice tablet -startdate now -boot c -drive file=/dev/migrate/mike_rhel6U1,if=virtio,format=qcow2,werror=stop,cache=none,boot=on -net nic,macaddr=54:52:40:3d:4c:58,vlan=0,model=virtio -net tap,script=/etc/qemu-ifup,downscript=no,vlan=0 -usb -spice ic=on,host=0,port=5910,disable-ticketing -qxl 1 -k en-us -balloon virtio -notify all
2.Start listenning port on Host t2
3.start #netserver on host t1
4.in the guest ,start 10 netperf client
ip=t1
i=0
for ((i=0;i<10;i++))
do
netperf -H $ip -l 3000 &
echo finish test_No. $i
done
5.do ping-pong live migration between 2 hosts
  
Actual results:
guest network down.
#dmesg in the guest
net eth0:Unexpected full queue
net eth0:Unexpected full queue
net eth0:Unexpected full queue
net eth0:Unexpected full queue
net eth0:Unexpected full queue
net eth0:Unexpected full queue

Expected results:


Additional info:
#service Network restart in the guest ,still can not get ipaddress

Comment 2 Mike Cao 2011-05-10 02:28:06 UTC
Tried start 10 netperf client in the guest w/o migration ,guest's network works
fine.

Comment 4 Mike Cao 2011-05-25 07:45:05 UTC
Tried RHEL5.7 32 bit guest (kernel 2.8.18-262.el5PAE) on kvm-83-235.el5

Hit the same issue.

Comment 8 Ronen Hod 2011-09-12 14:39:29 UTC
Hi Mike,

Since this is not reported by a customer, and we are very aggressive with closing RHEL5.8 bugs, and this looks like yet another high-load problem in RHEL5, I would close it.
Otherwise, mst had some questions in the previous comments.
Do you know if this problem exists in 6.2?

Comment 9 Mike Cao 2011-09-15 07:44:58 UTC
Hi,juzhang

Could you share the migration test results for RHEL6.2 whether hit this issue ?

Best Regards,
Mike

Comment 10 juzhang 2011-09-15 08:15:21 UTC
(In reply to comment #9)
> Hi,juzhang
> 
> Could you share the migration test results for RHEL6.2 whether hit this issue ?
> 
> Best Regards,
> Mike

Didn't hit this issue on rhel6.2.

Comment 13 RHEL Program Management 2012-04-02 10:45:45 UTC
This request was evaluated by Red Hat Product Management for inclusion
in a Red Hat Enterprise Linux release.  Product Management has
requested further review of this request by Red Hat Engineering, for
potential inclusion in a Red Hat Enterprise Linux release for currently
deployed products.  This request is not yet committed for inclusion in
a release.


Note You need to log in before you can comment on or make changes to this bug.