It is possible to emulate qMp with Virtual Box.

First of all open your virtualbox (VB) and create a new machine (type linux 2.6).
If you are using the VB wizard, do not select any storage device, we will do it latter.


To make the things easy it is better to use IDE instead of SATA.
Create a new IDE disk and browse between your foldersand select the qMp vdi image file.

The vdi file can be obtained here: with the name VBox-qMp_testing-factory-XXXXX.bin
It should be renamed to something such as VBox-qMp1.vdi

Network Configuration

First of all we create a virtual tun/tap device in the host machine.

ip tuntap add mode tap dev vbox 
ip link set vbox up
ip addr add dev vbox

In VB machine configuration at least one network interface must be added.
It should be configured as bridged interface to our new "vbox" network device.
The driver "PCnet-PCI II" works fine.


A serial port is needed because OpenWRT expects it. So it should be added in the VB machine configuration.

It is possible to use it to connect with the qMp node from the host machine (useful for debug), but it is NOT NEEDED.
To do so you should configure the VB serial as host pipe.
Check the option "create pipe" and add the path "/tmp/vboxS0".
Then in the host machine execute:

socat UNIX-CONNECT:/tmp/vboxS0 TCP-LISTEN:7777 &
netcat localhost 7777

And you will get serial!

Start VM

At this point you can start your virtual machine with qMp, open a browser and go to

If after the first reboot (automatic reboot) it does not boot (stays in "GRUB LOADING") reboot manually the VM.


Ok, but just one node is quite boring so let's go to create more of them. Just switch-off the virtual machine, right click and Clone it.
Select the option "regenerate MAC addresses", boot the new machine and then execute:

rm /qmp_configured
/etc/init.d/qmp_autoconf start

Then both nodes should see each other!


It is possible to use the same VirtualBox VDI image file. So again, you can get it from
Copy it to somewhere and change the name to, for instance "qMp1.vdi".

Then just start QEMU:

sudo qemu -net nic,model=pcnet -m 64 -hda qMp1.vdi

You can add "-nographic" option and "-daemonize".


Get serial is very easy, just add this option to QEMU

-serial pty

The qmeu will show you the pty device attached to the serial port of the virtual machine, then just use it:

screen /dev/pts/3



There are several ways to get the network working, I use the combo bridge+tap.
In the host machine we will have a bridge named "qemu" and the network interface of the virtual machine
will be a tap device attached to such bridge.

The next options must be added to the qemu exec command line

-net tap,ifname=qmp1

And these two scripts are needed:


if ! brctl show | grep qemu; then
    brctl addbr qemu
    ip link set qemu up
echo "Executing /etc/qemu-ifup" 
echo "Bringing up $1 for bridged mode..." 
sudo /sbin/ip link set $1 up promisc on
echo "Adding $1 to br0..." 
sudo brctl addif qemu $1
sleep 2


echo "Executing /etc/qemu-ifdown" 
sudo /sbin/ip link set $1 down
sudo /usr/sbin/brctl delif qemu $1
sudo /sbin/ip link delete dev $1

And make them executable

chmod +x /etc/qemu-*

To reach the node it is needed to add an IP to the host machine bridge:

ip addr add dev qemu

Multiple nodes

Each virtual machine needs its own disk file, so copy the VirtualBox vdi image to, for instance "qMpv2.vdi".
You should use a new one, do not copy the image from another existing machine!

To skip the problem of duplicate MAC addresses, the option "macaddr" should be added to the previous net option.

-net tap,ifname=qmp1,macaddr=$(/etc/qemu-mac)

And another script is needed (make it executable too!)


# generate a random mac address for the qemu nic
printf 'CA:FE:BA:BE:%02X:%02X\n' $((RANDOM%256)) $((RANDOM%256))

I use this little script to launch all the virtual machines from a directory:


for IMG in $VMDIR/*.vdi
    echo "Starting image $IMG" 
    sudo qemu -net nic,model=pcnet,macaddr=$(/etc/qemu-mac) \
    -net tap,ifname=qmp$(($RANDOM%30)) -m 64 -hda $IMG -nographic -serial pty -daemonize

Example of execution:

p4u@eni4c:~/virtual $ ls
qMp1.vdi  qMp2.vdi

p4u@eni4c:~/virtual $ sh 
Starting image ./qMp1.vdi
Executing /etc/qemu-ifup
Bringing up qmp4306 for bridged mode...
Adding qmp13 to br0...
char device redirected to /dev/pts/5

Starting image ./qMp2.vdi
Executing /etc/qemu-ifup
Bringing up qmp21308 for bridged mode...
Adding qmp21 to br0...
char device redirected to /dev/pts/7

A more advanced version of this scripts is attached as file to this article.

Updated by Roger Pueyo Centelles over 10 years ago ยท 18 revisions