Home  >  Article  >  Backend Development  >  Detailed explanation of the configuration xml code in the KVM virtual machine

Detailed explanation of the configuration xml code in the KVM virtual machine

Y2J
Y2JOriginal
2017-04-24 09:17:343220browse

In RHEL6, the XML file used to boot from disk

Here is dcs01.xml as an example:

 <domain type=&#39;kvm&#39;>
<name>dcs01</name>
<uuid>e5fff551-bbe1-e748-c8e4-8ecb3bffb902</uuid>
<memory>1048576</memory>
<currentMemory>1048576</currentMemory>
<vcpu>1</vcpu>
<os>
<type arch=&#39;x86_64&#39; machine=&#39;rhel6.0.0&#39;>hvm</type>
<boot dev=&#39;hd&#39;/>
</os>
<features>
<acpi/>
<apic/>
<pae/>
</features>
<clock offset=&#39;localtime&#39;/>
<on_poweroff>destroy</on_poweroff>
<on_reboot>restart</on_reboot>
<on_crash>restart</on_crash>
<devices>
<emulator>/usr/libexec/qemu-kvm</emulator>
<disk type=&#39;file&#39; device=&#39;disk&#39;>
<driver name=&#39;qemu&#39; type=&#39;raw&#39; cache=&#39;none&#39;/>
<source file=&#39;/home/kvm/images/dcs01.img&#39;/>
<target dev=&#39;hda&#39; bus=&#39;ide&#39;/>
<address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;0&#39; unit=&#39;0&#39;/>
</disk>
<disk type=&#39;file&#39; device=&#39;cdrom&#39;>
<driver name=&#39;qemu&#39; type=&#39;raw&#39;/>
<target dev=&#39;hdc&#39; bus=&#39;ide&#39;/>
<readonly/>
<address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;1&#39; unit=&#39;0&#39;/>
</disk>
<controller type=&#39;ide&#39; index=&#39;0&#39;>
<address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x01&#39; function=&#39;0x1&#39;/>
</controller>
<interface type=&#39;bridge&#39;>
<mac address=&#39;52:54:00:ad:75:98&#39;/>
<source bridge=&#39;br0&#39;/>
<address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x03&#39; function=&#39;0x0&#39;/>
</interface>
<input type=&#39;tablet&#39; bus=&#39;usb&#39;/>
<input type=&#39;mouse&#39; bus=&#39;ps2&#39;/>
<graphics type=&#39;vnc&#39; port=&#39;-1&#39; autoport=&#39;yes&#39;/>
<video>
<model type=&#39;vga&#39; vram=&#39;9216&#39; heads=&#39;1&#39;/>
<address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x02&#39; function=&#39;0x0&#39;/>
</video>
<memballoon model=&#39;virtio&#39;>
<address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x05&#39; function=&#39;0x0&#39;/>
</memballoon>
</devices>
</domain>

Comments are as follows:

 1. Kvm guest definition starts

<domain type=&#39;kvm&#39;>

 2. Short name of guest. It consists of letters and numbers and cannot contain spaces

<name>dcs01</name>

 3. uuid, generated by the command line tool uuidgen.

<uuid>e5fff551-bbe1-e748-c8e4-8ecb3bffb902</uuid>

4. The maximum memory that the guest can use without rebooting the guest, in KB.

<memory>1048576</memory>

5. The memory when the guest starts can be passed Use virsh setmem to adjust the memory, but it cannot be larger than the maximum usable memory.

<currentMemory>1048576</currentMemory>

6. Assigned virtual cpu

<vcpu>1</vcpu>

7. Related OS
architecture: i686, x86_64
machine: host operating system
boot: Specify the boot device. You can repeat multiple lines and specify different values ​​as a boot device list.

<os>
<type arch=&#39;x86_64&#39; machine=&#39;rhel6.0.0&#39;>hvm</type>
<boot dev=&#39;hd&#39;/>
</os>

8. Processor features

<features>
<acpi/>
<apic/>
<pae/>
</features>

9. Clock. Use local time: localtime

<clock offset=&#39;localtime&#39;/>

10. Define the default actions when power off, reboot, or crash in the kvm environment are destroy and restart respectively. Other allowed actions include: preserve, rename-restart.
destroy: Stop the virtual machine. Equivalent to turning off the power.
restartRestart the virtual machine.

<on_poweroff>destroy</on_poweroff>
<on_reboot>restart</on_reboot>
<on_crash>restart</on_crash>

 11. Start of device definition

<devices>

 12. Simulation element, the writing method here is used for guest

<emulator>/usr/libexec/qemu-kvm</emulator>

 13. File used for kvm storage. In this example, it appears as an IDE device in the guest.
Use the qemu-img command to create this file. The default directory of kvm image is: /var/lib/libvirt/images/

<disk type=&#39;file&#39; device=&#39;disk&#39;>
<driver name=&#39;qemu&#39; type=&#39;raw&#39; cache=&#39;none&#39;/>
<source file=&#39;/home/kvm/images/dcs01.img&#39;/>
<target dev=&#39;hda&#39; bus=&#39;ide&#39;/>
<address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;0&#39; unit=&#39;0&#39;/>
</disk>

Supplement: Multiple disks can be defined.
Use virtio:
Use an ordinary driver, that is, when both the hard disk and the network card adopt the default configuration, the network card works under the simulated rtl 8139 network card, and the speed is 100M full duplex. After using the virtio driver, the network card works in 1000M mode.

Use an ordinary driver, that is, when both the hard disk and the network card adopt the default configuration, the hard disk is in IDE mode. After using the virtio driver, the hard disk works in SCSI mode.

<disk type=&#39;file&#39; device=&#39;disk&#39;>
<driver name=&#39;qemu&#39; type=&#39;raw&#39;/>
<source file=&#39;/usr/local/kvm/vmsample/disk.os&#39;/>
<target dev=&#39;vda&#39; bus=&#39;virtio&#39;/>
</disk>

CD-ROM device:

<disk type=&#39;file&#39; device=&#39;cdrom&#39;>
<driver name=&#39;qemu&#39; type=&#39;raw&#39;/>
<target dev=&#39;hdc&#39; bus=&#39;ide&#39;/>
<readonly/>
<address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;1&#39; unit=&#39;0&#39;/>
</disk>

14. Use the bridge type. Make sure each kvm guest's mac address is unique. A tun device will be created with the name vnetx (x is 0,1,2...)

<interface type=&#39;bridge&#39;>
<mac address=&#39;52:54:00:ad:75:98&#39;/>
<source bridge=&#39;br0&#39;/>
<address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x03&#39; function=&#39;0x0&#39;/>
</interface>

Supplement:
Use the default virtual network instead of the bridge, that is, the guest is in NAT mode. You can also omit the mac address element, so the mac address will be automatically generated.

<interface type=&#39;network&#39;>
<source network=&#39;default&#39;/>
<mac address="3B:6E:01:69:3A:11"/>
</interface>

The address 192.168.122.x/24 is assigned by default, and can also be specified manually. The gateway is 192.168.122.1

Use virtio:
Use an ordinary driver, that is, when the hard disk and network card are both configured in the default configuration, the network card works under the simulated rtl 8139 network card, with a speed of 100M. duplex. After using the virtio driver, the network card works in 1000M mode.

<interface type=&#39;bridge&#39;>
<source bridge=&#39;br1&#39;/>
<model type=&#39;virtio&#39; />
</interface>

15. Input device

<input type=&#39;tablet&#39; bus=&#39;usb&#39;/>
<input type=&#39;mouse&#39; bus=&#39;ps2&#39;/>

16. Define the graphics device that interacts with guset. In this example, the vnc protocol is used. The address of listen is the address of host. prot is -1, which means the port number is automatically assigned. Use the following command to find the port number:
virsh vncdisplay 33e1e9ce7395ccd8e6281087cf15dba5

It is not set here

<graphics type=&#39;vnc&#39; port=&#39;-1&#39; autoport=&#39;yes&#39;/>

 17 , Device definition ends

</devices>

18. KVM definition ends

</domain>

centos_x86_6.4
b9dcdd92-9b9b-14d6-3938-1982a9746a12
2097152
2097152
1
hvm
destroy
restart
restart
/bin/qemu-kvm

  <disk type=&#39;file&#39; device=&#39;disk&#39;>
      <driver name=&#39;qemu&#39; type=&#39;qcow2&#39;/>

#The destination mirror path in this example, Shown as IDE device in guest.​

<source file=&#39;/home/template_make/centos_x86_6.4.img&#39;>
        <seclabel model=&#39;selinux&#39; relabel=&#39;no&#39;/>
      </source>
      <target dev=&#39;hda&#39; bus=&#39;ide&#39;/>
      <alias name=&#39;ide0-0-0&#39;/>
      <address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;0&#39; target=&#39;0&#39; unit=&#39;0&#39;/>
    </disk>
    <disk type=&#39;file&#39; device=&#39;cdrom&#39;>
      <driver name=&#39;qemu&#39; type=&#39;raw&#39;/>
      <source file=&#39;/home/template_make/CentOS-6.4-x86_64-bin-DVD1.iso&#39;/>
      <target dev=&#39;hdc&#39; bus=&#39;ide&#39;/>
      <readonly/>
      <alias name=&#39;ide0-1-0&#39;/>
      <address type=&#39;drive&#39; controller=&#39;0&#39; bus=&#39;1&#39; target=&#39;0&#39; unit=&#39;0&#39;/>
    </disk>
    <controller type=&#39;usb&#39; index=&#39;0&#39;>
      <alias name=&#39;usb0&#39;/>
      <address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x01&#39; function=&#39;0x2&#39;/>
    </controller>
    <controller type=&#39;ide&#39; index=&#39;0&#39;>
      <alias name=&#39;ide0&#39;/>
      <address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x01&#39; function=&#39;0x1&#39;/>
    </controller>
    <interface type=&#39;bridge&#39;>

#Virtual machine network connection method

 <mac address=&#39;52:54:00:78:f9:5a&#39;/>
      <source bridge=&#39;br0&#39;/>
      <target dev=&#39;vnet27&#39;/>

## Use virtio: Use an ordinary driver, that is, when both the hard disk and the network card adopt the default configuration, the hard disk is in ide mode, and the network card works under the simulated rtl 8139 network card, with a speed of 100M full duplex . After using the virtio driver, the network card works in 1000M mode, and the hard disk works in SCSI mode.

<model type=&#39;virtio&#39;/>
      <alias name=&#39;net0&#39;/>
      <address type=&#39;pci&#39; domain=&#39;0x0000&#39; bus=&#39;0x00&#39; slot=&#39;0x03&#39; function=&#39;0x0&#39;/>
    </interface>
    <input type=&#39;mouse&#39; bus=&#39;ps2&#39;/>

#Log in with vnc, the port number is automatically assigned. You can query [vncdisplay domainId] through virsh vncdisplay


      
    
    


The above is the detailed content of Detailed explanation of the configuration xml code in the KVM virtual machine. For more information, please follow other related articles on the PHP Chinese website!

Statement:
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn