Bug 680047

Summary: Kernel panic after starting an LXC guest with veth NICs / CLONE_NEWNET
Product: Red Hat Enterprise Linux 6 Reporter: wangyimiao <yimwang>
Component: kernelAssignee: Thomas Graf <tgraf>
Status: CLOSED DUPLICATE QA Contact: Red Hat Kernel QE team <kernel-qe>
Severity: high Docs Contact:
Priority: high    
Version: 6.1CC: berrange, dyuan, eblake, gren, llim, rkhan, vbian, xen-maint
Target Milestone: rcKeywords: Regression
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-03-08 20:38:22 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
Kernel panic seen when launching a container
none
host crash&restart message log none

Description wangyimiao 2011-02-24 08:33:05 UTC
Description of problem:
Host OS will be restart after start LXC domain guest

Version-Release number of selected component (if applicable):
libvirt-0.8.7-7.el6.x86_64
qemu-kvm-0.12.1.2-2.145.el6.x86_64
qemu-img-0.12.1.2-2.145.el6.x86_64
kernel-2.6.32-113.el6.x86_64

How reproducible:
5/5

Steps to Reproduce:
1.# cat vm1.xml 
<domain type='lxc'>
  <name>vm1</name>
  <uuid>386f5b25-43ee-9d62-4ce2-58c3809e47c1</uuid>
  <memory>500000</memory>
  <currentMemory>500000</currentMemory>
  <vcpu>1</vcpu>
  <os>
    <type arch='x86_64'>exe</type>
    <init>/bin/sh</init>
  </os>
  <clock offset='utc'/>
  <on_poweroff>destroy</on_poweroff>
  <on_reboot>restart</on_reboot>
  <on_crash>destroy</on_crash>
  <devices>
    <emulator>/usr/libexec/libvirt_lxc</emulator>
    <interface type='network'>
      <mac address='52:54:00:f2:2c:ac'/>
      <source network='default'/>
      <target dev='veth0'/>
    </interface>
    <console type='pty'>
      <target port='0'/>
    </console>
  </devices>
</domain>
2.Defined and start the LXC domain guest.

# virsh -c lxc:/// define  vm1.xml 
Domain vm1 defined from vm1.xml

# virsh -c lxc:/// start vm1 
Domain vm1 started

After that host OS will be restart.
  
Actual results:
Host OS will be restart after start LXC domain guest

Expected results:
After start LXC domain guest,host OS should be works fine.

Additional info:
Note:If I change kernel to verison :(kernel-2.6.32-71.17.1.el6.x86_64),host os will works fine after start LXC domain guest.

Comment 1 Daniel Berrangé 2011-02-24 09:57:16 UTC
Can you describe what happens to the host OS in more details. Are you getting a kernel panic ? Or is it an instant reboot ? Or does it shutdown gracefully & reboot ?

Comment 2 Daniel Berrangé 2011-02-24 10:23:08 UTC
Created attachment 480707 [details]
Kernel panic seen when launching a container

Comment 3 Daniel Berrangé 2011-02-24 11:05:59 UTC
Further points of reference:

 - 2.6.32-109.el6.x86_64  - crashes
 - 2.6.32-94.el6.x86_64   - works fine

The crash can be avoided by removing the <interface> element from the XML for the guest. The panic in comment #2 also includes networking functions in the stack trace. So these two things point to a regression in the networking stack vs 6.0 kernels

Comment 5 wangyimiao 2011-02-24 11:22:29 UTC
Created attachment 480723 [details]
host crash&restart message log

The host os first crash and then will be restart.
The details message please sees the attachment.

Comment 6 Vivian Bian 2011-02-25 06:10:41 UTC
tested with 
libvirt-0.8.7-8.el6.x86_64
qemu-kvm-0.12.1.2-2.147.el6.x86_64
kernel-2.6.32-117.el6.x86_64


Steps to Reproduce:
1.# cat vm1.xml 
<domain type='lxc'>
  <name>vm1</name>
  <uuid>386f5b25-43ee-9d62-4ce2-58c3809e47c1</uuid>
  <memory>500000</memory>
  <currentMemory>500000</currentMemory>
  <vcpu>1</vcpu>
  <os>
    <type arch='x86_64'>exe</type>
    <init>/bin/sh</init>
  </os>
  <clock offset='utc'/>
  <on_poweroff>destroy</on_poweroff>
  <on_reboot>restart</on_reboot>
  <on_crash>destroy</on_crash>
  <devices>
    <emulator>/usr/libexec/libvirt_lxc</emulator>
    <interface type='network'>
      <mac address='52:54:00:f2:2c:ac'/>
      <source network='default'/>
      <target dev='veth0'/>
    </interface>
    <console type='pty'>
      <target port='0'/>
    </console>
  </devices>
</domain>
2.Defined and start the LXC domain guest.

# virsh -c lxc:/// define  vm1.xml 
Domain vm1 defined from vm1.xml

# virsh -c lxc:/// start vm1 
Domain vm1 started


Tried 5 times , and kernel panic twice . So with the new version the reproducible percentage is not 100%

Comment 7 Thomas Graf 2011-03-08 20:38:22 UTC

*** This bug has been marked as a duplicate of bug 677786 ***