[one-users] Error on start-up

Rich Wellner rkw at objenv.com
Thu Sep 30 13:13:26 PDT 2010


  Hi,

I'm trying to launch my first KVM machine via one and am running into 
what looks like a network problem.

 From one, the error log is:

/Wed Sep 29 22:24:26 2010 [LCM][I]: New VM state is BOOT
Wed Sep 29 22:24:26 2010 [VMM][I]: Generating deployment file: 
/var/lib/one/12/deployment.0
Wed Sep 29 22:24:56 2010 [VMM][I]: Command execution fail: 'mkdir -p 
/home/sharedFS/images/12/images && cat > 
/home/sharedFS/images/12/images/deployment.0 && virsh --connect 
qemu:///system create /home/sharedFS/images/12/images/deployment.0'
Wed Sep 29 22:24:56 2010 [VMM][I]: STDERR follows.
Wed Sep 29 22:24:56 2010 [VMM][I]: error: Failed to create domain from 
/home/sharedFS/images/12/images/deployment.0
Wed Sep 29 22:24:56 2010 [VMM][I]: error: monitor socket did not show 
up.: Connection refused
Wed Sep 29 22:24:56 2010 [VMM][I]: ExitCode: 1
Wed Sep 29 22:24:56 2010 [VMM][E]: Error deploying virtual machine
Wed Sep 29 22:24:56 2010 [DiM][I]: New VM state is FAILED/

I trace that down to the worker node.  I find a log file 
/var/log/libvirt/qemu/one-12.log containing:

/LC_ALL=C 
PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/sbin:/sbin:/bin 
/usr/bin/kvm -S -M pc-0.12 -enable-kvm -m 128 -smp 1 -name one-12 -uuid 
1cd58679-9171-ad5e-f22c-8df0cc6f7f16 -nographic -chardev 
socket,id=monitor,path=/var/lib/libvirt/qemu/one-12.monitor,server,nowait -monitor 
chardev:monitor -boot c -kernel /vmlinuz -initrd /initrd.img -drive 
file=/home/sharedFS/images/12/images/disk.0,if=ide,index=0,boot=on -net 
nic,macaddr=00:2a:f9:42:1e:f7,vlan=0,name=nic.0 -net 
tap,vlan=0,name=tap.0 -serial none -parallel none -usb
warning: could not configure /dev/net/tun: no virtual network emulation/

Since the one error sounds network related I start to focus on that 
warning, even though it's just a warning.  So I try running the machine 
from a command line:

/$ sudo qemu-system-x86_64 -k en-us -vga vmware -hda 100924-1.img -net 
nic,macaddr=`macgen` -net tap -boot c -m 384
+ switch=br0
+ [ -n tap4 ]
+ whoami
+ /usr/bin/sudo /usr/sbin/tunctl -u root -t tap4
TUNSETIFF: Device or resource busy
+ /usr/bin/sudo /sbin/ip link set tap4 up
+ sleep 0.5s
+ /usr/bin/sudo /usr/sbin/brctl addif br0 tap4
+ exit 0
/
The interesting part is "TUNSETIFF: Device or resource busy".  I say 
it's interesting because, again, it seems network related and further, 
if I run those commands from a shell, they work just fine.

/$ /usr/bin/sudo /usr/sbin/tunctl -u root -t tap4
Set 'tap4' persistent and owned by uid 0
/
So I'm not sure why when being launched by qemu/one I get that error.  
And I'm not even sure it's the root cause.  Finally, when I launch the 
VM manually (i.e. without one, just with qemu) even though that 'busy' 
error shows up the machine starts and networking is completely functional.

Any ideas?  What else do I need to provide?

rw2

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20100930/d760620d/attachment-0001.htm>


More information about the Users mailing list