kvm

content history

15y, 298d ago [edited]

KVM guest management

Install packages

Debian or Ubuntu

sudo apt-get install qemu-kvm libvirt-bin ubuntu-vm-builder bridge-utils

Redhat or Fedora

yum install kvm
yum install virt-manager libvirt libvirt-python python-virtinst
chkconfig libvirtd on
service libvirtd start

KVM virsh cheat sheet!

This is a list of commands to use in the virsh cli.

To start the virsh cli run the following cmd:

$ virsh

or:

virsh --connect qemu:///system

KVM has its own commands, but we like to use virsh.

Here is a list comparing Xen and KVM (virsh) commands.

List running guests

xen: xm list
kvm: virsh list
virsh dominfo <dom name>

Start a guest

xen: xm create /etc/xen/<dom name>
kvm: virsh create /etc/libvirt/qemu/<dom name>.xml
virsh start <dom name>

Console to guest

xen: xm console <dom name>
kvm: virsh console <dom name>

Shutdown a guest

xen: xm shutdown <dom name>
kvm: virsh shutdown <dom name>

This command doesn't always work...

Be sure the guests XML file has the following entry:

<feature>
<acpi/>
</feature>

Also make sure the apci is installed on the guest operating system.

debian / ubuntu: sudo apt-get install acpid
fedora / redhat: yum install acpid

Autostart guest on reboot

xen: ln -s /etc/xen/<dom name> /etc/xen/auto/<dom name>
kvm: virsh autostart <dom name>

Edit a guests configuration file

xen: NA
kvm: virsh edit <dom name>

Pull the power on a guest

xen: xm destroy <dom name>
kvm: virsh destroy <dom name>

Quit virsh cli

kvm: virsh quit

Virtual Serial Console configuration

From the KVM host run:

virsh edit <dom>

Then add the following inside the <device></device> directives:

<serial type="pty">
    <source path="/dev/pts/3" />
    <target port="1" />
</serial>

From the Guest run:

sudo cp -p /etc/init/tty06.conf /etc/init/ttyS0.conf
sudo vi /etc/init/ttyS0.conf
sudo diff /etc/init/tty06.conf /etc/init/ttyS0.conf

Vmbuilder Example

This example assumes ubuntu linux.

Install vmbuilder

sudo apt-get install python-vm-builder

Display Help vmbuilder

vmbuilder kvm ubuntu --help

Custom vmbuilder string

mbison example

VMNAME=mbison          
mkdir /vms/$VMNAME         

sudo vmbuilder kvm ubuntu --libvirt=qemu:///system --suite=lucid --flavour=virtual \           
--hostname=$VMNAME --domain='foxhop.net' --rootsize='10240' --mem='1024' \         
--ip=192.168.1.51 --gw=192.168.1.254 --dns=192.168.1.22 --bridge=br0 \         
--addpkg=openssh-server --addpkg=acpid --timezone=EDT --verbose

cammy example

VMNAME=cammy           
mkdir /vms/$VMNAME         

sudo vmbuilder kvm ubuntu       \          
    --libvirt=qemu:///system    \          
    --suite=lucid               \          
    --arch=amd64 -o             \          
    --flavour=virtual           \          
    --hostname=$VMNAME          \          
    --dest=/vms/$VMNAME         \          
    --rootsize=20480            \          
    --mem=1024                  \          
    --bridge=br0                \          
    --ip=192.168.1.52           \          
    --gw=192.168.1.254          \          
    --dns=192.168.1.22          \          
    --user=john                 \          
    --pass=doe                  \          
    --addpkg=openssh-server     \          
    --addpkg=acpid              \          
    --timezone=EDT              \          
    --verbose

virt-install

debian netboot example

This method shows the virt-install script installing from a debian netboot image hosted on the internet

HOSTNAME=tehforum
DOMAIN=foxhop.net

sudo virt-install \
--name=$HOSTNAME \
--vcpu=1 \
--ram=396 \
--disk=/KVMROOT/$HOSTNAME.qcow2,size=10 \
--os-type=linux \
--autostart \
--location=http://ftp.nl.debian.org/debian/dists/wheezy/main/installer-amd64/ \
--extra-args="auto=true priority=critical keymap=us locale=en_US hostname=$HOSTNAME domain=$DOMAIN url=http://192.168.1.22/foxhop-debconf-preseed.txt"

ubuntu netboot example

This method shows the virt-install script installing from an ubuntu netboot image hosted on the internet

HOSTNAME=mbison
DOMAIN=foxhop.net

sudo virt-install \
--name=$HOSTNAME \
--vcpu=1 \
--ram=396 \
--disk=/KVMROOT/$HOSTNAME.qcow2,size=10 \
--os-type=linux \
--autostart \
--location=http://archive.ubuntu.com/ubuntu/dists/raring/main/installer-amd64/ \
--extra-args="auto=true priority=critical keymap=us locale=en_US hostname=$HOSTNAME domain=$DOMAIN url=http://192.168.1.22/foxhop-debconf-preseed.txt"

Mounting a qcow2 image on the host

Sometimes it is helpful to be able to mount a drive image under the host system. For example, if the guest doesn't have network support, the only way to transfer files into and out of the guest will be by the storage devices it can address or to restore files from a backup image.

To mount qcow2 images there is (at least in F-11 qemu) very useful qemu-nbd util. It shares image through kernel network block device protocol and this allows to mount it:

sudo modprobe nbd max_part=63
sudo qemu-nbd -c /dev/nbd0 image.img
sudo mount /dev/nbd0p1 /mnt/image

Add another disk image to a guest

This is how you add an additional disk to a guest. In this example I will be adding a 20GB disk image to a guest dom named cammy.

fallocate -l 20g /vms/cammy/cammy-aux.img

qemu-img create -f qcow2 /vms/cammy/cammy-aux.img 20G

If you want to look that the image statistics try this command.

qemu-img info /vms/cammy/cammy-aux.img

Now we need to mount this disk:

sudo virsh attach-disk cammy /vms/cammy/cammy-aux.img vdb

That command only mounted the disk. To make this change permanent we need to alter the dom's xml:

# sudo virsh edit cammy

 <disk type='file' device='disk'>
   <driver name='qemu' type='qcow2'/>
   <source file='/KVMROOT/cammy.qcow2'/>
   <target dev='vda' bus='virtio'/>
   <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/>
 </disk>

 <disk type='file' device='disk'>
   <driver name='qemu' type='raw'/>
   <source file='/KVMROOT/cammy-aux.img'/>
   <target dev='vdb' bus='virtio'/>
 <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/>

SSH to the guest and verify that the disk shows up in fdisk -l. You may now partition this disk how you like. For more information view linux raw filesystem management.

shrink a qcow2 image

http://pve.proxmox.com/wiki/Shrink_Qcow2_Disk_Files

KVM guest management
=======================
.. contents::

Install packages 
===============================
**Debian or Ubuntu**

* sudo apt-get install qemu-kvm libvirt-bin ubuntu-vm-builder bridge-utils

**Redhat or Fedora**

* yum install kvm 
* yum install virt-manager libvirt libvirt-python python-virtinst 
* chkconfig libvirtd on 
* service libvirtd start

KVM virsh cheat sheet!
============================
This is a list of commands to use in the virsh cli.

To start the virsh cli run the following cmd:

.. code-block:: bash
  
  $ virsh

or:

.. code-block:: bash

virsh --connect qemu:///system

KVM has its own commands, but we like to use virsh.

Here is a list comparing Xen and KVM (virsh) commands.

List running guests 
---------------------------------------------- 
**xen** 
 | xm list

**kvm** 
 | virsh list 
 | virsh dominfo <dom name>

Start a guest 
---------------------------------------------- 
**xen** 
 | xm create /etc/xen/<dom name>

**kvm** 
 | virsh create /etc/libvirt/qemu/<dom name>.xml
 | virsh start <dom name>

Console to guest
---------------------------------------------- 
**xen** 
 xm console <dom name>

**kvm** 
 virsh console <dom name>

Shutdown a guest 
---------------------------------------------- 
**xen** 
 xm shutdown <dom name>

**kvm** 
 virsh shutdown <dom name>

This command doesn't always work...

Be sure the guests XML file has the following entry:

| <feature> 
 |  <acpi/>
 | </feature>

Also make sure the apci is installed on the guest operating system.

**debian / ubuntu**
 sudo apt-get install acpid

**fedora / redhat**
 yum install acpid

Autostart guest on reboot 
---------------------------------------------- 
**xen** 
 ln -s /etc/xen/<dom name> /etc/xen/auto/<dom name>

**kvm** 
 virsh autostart <dom name>

Edit a guests configuration file
-----------------------------------
**xen**
  NA

**kvm**  
 virsh edit <dom name>

Pull the power on a guest
----------------------------
**xen**
 xm destroy <dom name>

**kvm**
 virsh destroy <dom name>

Quit virsh cli
-----------------
**kvm** 
 virsh quit

Virtual Serial Console configuration
======================================
**From the KVM host run:**

.. code-block:: xml

virsh edit <dom>

Then add the following inside the <device></device> directives:

.. code-block:: xml

**From the Guest run:**

.. code-block:: xml

sudo cp -p /etc/init/tty06.conf /etc/init/ttyS0.conf
 sudo vi /etc/init/ttyS0.conf
 sudo diff /etc/init/tty06.conf /etc/init/ttyS0.conf

Vmbuilder Example
======================
This example assumes ubuntu linux.

Install vmbuilder
---------------------
.. code-block:: python

sudo apt-get install python-vm-builder

Display Help vmbuilder
--------------------------------
.. code-block:: python

vmbuilder kvm ubuntu --help

Custom vmbuilder string	
-------------------------------			
 			
mbison example

.. code-block:: bash			
		
 VMNAME=mbison			
 mkdir /vms/$VMNAME			
 			
 sudo vmbuilder kvm ubuntu --libvirt=qemu:///system --suite=lucid --flavour=virtual \			
 --hostname=$VMNAME --domain='foxhop.net' --rootsize='10240' --mem='1024' \			
 --ip=192.168.1.51 --gw=192.168.1.254 --dns=192.168.1.22 --bridge=br0 \			
 --addpkg=openssh-server --addpkg=acpid --timezone=EDT --verbose			
	 			
cammy example			
 			
.. code-block:: bash			
 			
 VMNAME=cammy			
 mkdir /vms/$VMNAME			
 			
 sudo vmbuilder kvm ubuntu       \			
     --libvirt=qemu:///system    \			
     --suite=lucid               \			
     --arch=amd64 -o             \			
     --flavour=virtual           \			
     --hostname=$VMNAME          \			
     --dest=/vms/$VMNAME         \			
     --rootsize=20480            \			
     --mem=1024                  \			
     --bridge=br0                \			
     --ip=192.168.1.52           \			
     --gw=192.168.1.254          \			
     --dns=192.168.1.22          \			
     --user=john                 \			
     --pass=doe                  \			
     --addpkg=openssh-server     \			
     --addpkg=acpid              \			
     --timezone=EDT              \ 			
     --verbose

virt-install
================
debian netboot example
---------------------------

This method shows the virt-install script installing from a debian netboot image hosted on the internet

.. code-block:: bash

HOSTNAME=tehforum
 DOMAIN=foxhop.net

sudo virt-install \
 --name=$HOSTNAME \
 --vcpu=1 \
 --ram=396 \
 --disk=/KVMROOT/$HOSTNAME.qcow2,size=10 \
 --os-type=linux \
 --autostart \
 --location=http://ftp.nl.debian.org/debian/dists/wheezy/main/installer-amd64/ \
 --extra-args="auto=true priority=critical keymap=us locale=en_US hostname=$HOSTNAME domain=$DOMAIN url=http://192.168.1.22/foxhop-debconf-preseed.txt"

ubuntu netboot example
----------------------------

This method shows the virt-install script installing from an ubuntu netboot image hosted on the internet

.. code-block:: bash

HOSTNAME=mbison
 DOMAIN=foxhop.net

sudo virt-install \
 --name=$HOSTNAME \
 --vcpu=1 \
 --ram=396 \
 --disk=/KVMROOT/$HOSTNAME.qcow2,size=10 \
 --os-type=linux \
 --autostart \
 --location=http://archive.ubuntu.com/ubuntu/dists/raring/main/installer-amd64/ \
 --extra-args="auto=true priority=critical keymap=us locale=en_US hostname=$HOSTNAME domain=$DOMAIN url=http://192.168.1.22/foxhop-debconf-preseed.txt"

Mounting a qcow2 image on the host
======================================

To mount qcow2 images there is (at least in F-11 qemu) very useful qemu-nbd util. It shares image through kernel network block device protocol and this allows to mount it:

.. code-block:: bash
   
  sudo modprobe nbd max_part=63
  sudo qemu-nbd -c /dev/nbd0 image.img
  sudo mount /dev/nbd0p1 /mnt/image

Add another disk image to a guest
====================================

This is how you add an additional disk to a guest.  In this example I will be adding a 20GB disk image to a guest dom named cammy.

.. code-block:: bash

fallocate -l 20g /vms/cammy/cammy-aux.img

.. code-block:: bash

qemu-img create -f qcow2 /vms/cammy/cammy-aux.img 20G

If you want to look that the image statistics try this command.

.. code-block:: bash

qemu-img info /vms/cammy/cammy-aux.img

Now we need to mount this disk:

.. code-block:: bash

sudo virsh attach-disk cammy /vms/cammy/cammy-aux.img vdb

That command only mounted the disk.  To make this change permanent we need to alter the dom's xml:

.. code-block:: xml

# sudo virsh edit cammy

SSH to the guest and verify that the disk shows up in *fdisk -l*.  You may now partition this disk how you like.  For more information view `linux raw filesystem management <http://www.foxhop.net/linux-raw-filesystem-management>`_.

shrink a qcow2 image
========================

http://pve.proxmox.com/wiki/Shrink_Qcow2_Disk_Files

Comments

Leave first comment to start a conversation!