Virtio

So-called "full virtualization" is a nice feature because it allows you to run any operating system virtualized. However, it's slow because the hypervisor has to emulate actual physical devices such as RTL8139 network cards . This emulation is both complicated and inefficient.

Virtio is a virtualization standard for network and disk device drivers where just the guest's device driver "knows" it is running in a virtual environment, and cooperates with the hypervisor. This enables guests to get high performance network and disk operations, and gives most of the performance benefits of paravirtualization.

Note that virtio is different, but architecturally similar to, Xen paravirtualized device drivers (such as the ones that you can install in a Windows guest to make it go faster under Xen). Also similar is VMWare's Guest Tools.

This page describes how to configure libvirt to use virtio with KVM guests.

Requirements

  • KVM or recent (not 0.9.1) development QEMU

  • A virtio-compatible guest: any Linux OS with kernel >= 2.6.25 should be OK. Fedora 9 and above are explicitly supported.

  • libvirt >= 0.4.4

We assume that you have installed the virtio-compatible guest under KVM using libvirt (ie. using something like virt-install or virt-manager).

For a Windows guest, it is necessary to install the latest virtio drivers into the guest OS. Additional network and disk tips are available, outlining the process for installing drivers.

Network driver

First, shut down the guest and then edit its configuration file:

virsh edit guestname

In the <interface> section, add a virtio model, like this:

<interface type='network'>
  ...
  <model type='virtio' />
</interface>

When you boot the guest (virsh start guestname), if it worked you should still have a working network, and you should see (from inside the guest) that you are using the virtio_net driver:

# /sbin/lsmod | grep virtio
[shows virtio_pci, virtio_net and others loaded]
# cat /sys/devices/virtio-pci/0/net/eth0/statistics/rx_bytes
...

If it doesn't work, then check the following file in the host for errors:

/var/log/libvirt/qemu/[guestname].log

There are quite a lot of things that could go wrong such as: not using KVM, or not using a sufficiently recent version of KVM.

Disk (block) device driver

Similar to above, except the the configuration file should be changed to e.g.:

<disk type='...' device='disk'>
  ...
  <target dev='vda' bus='virtio'/>
</disk>

If there remove any

<address .../>

element for this disk that may exist, allowing libvirt to regenerated it appropriately.

However, if this is to be the disk which holds the guest's root filesystem, you first need to ensure that the guest will be able to mount the virtio disk during bootup.

On Fedora 9 or later, you can do this using mkinitrd:

# mkinitrd --with virtio_pci --with virtio_blk -f /boot/initrd-$(uname -r).img $(uname -r)

Note, this step is only needed in order to transition a guest from IDE or SCSI to virtio. If you initially install the guest using a virtio disk, or if you update the kernel package while booted from a virtio disk, then this step is not needed.