From KVM
Line 2: Line 2:
 
* Virtio was chosen to be the main platform for IO virtualization in KVM
 
* Virtio was chosen to be the main platform for IO virtualization in KVM
 
* The idea behind it is to have a common framework for hypervisors for IO virtualization
 
* The idea behind it is to have a common framework for hypervisors for IO virtualization
* More information (although not uptodate) can be found in [[http://kvm.qumranet.com/kvmwiki/KvmForum2007?action=[[AttachFile]]&do=get&target=kvm_pv_drv.pdf]]
+
* More information (although not uptodate) can be found in kvm [http://kvm.qumranet.com/kvmwiki/KvmForum2007?action=[[AttachFile]]&do=get&target=kvm_pv_drv.pdf pv driver]  
 
* At the moment network/block/balloon devices are suported for kvm
 
* At the moment network/block/balloon devices are suported for kvm
* The host impelementation is in userspace - qemu, so no driver is needed in the host.
+
* The host implementation is in userspace - qemu, so no driver is needed in the host.
  
 
= How to use Virtio =
 
= How to use Virtio =
 
* Get kvm version >= 60
 
* Get kvm version >= 60
 
* Get Linux kernel with virtio drivers for the guest
 
* Get Linux kernel with virtio drivers for the guest
** Soon a link will be updated
+
** Either build it around Rusty's tree [http://ozlabs.org/~rusty/kernel/hg/ repo]
 +
** Or git clone git://kvm.qumranet.com/home/dor/src/linux-2.6-nv use branch rusty
 +
** Soon an official repository will be released
 
** As an alternative one can use a standard guest kernel for the guest > 2.6.18 and make use sync backward compatibility option
 
** As an alternative one can use a standard guest kernel for the guest > 2.6.18 and make use sync backward compatibility option
 +
** Backport and instructions can be found in Anthony Liguori's [http://codemonkey.ws/virtio-ext-modules virtio-ext-modules]
 +
** At the moment it's broken since the guest got developed, soon update
 
* Use model=virtio for the network devices.
 
* Use model=virtio for the network devices.
 +
** Example
 +
<pre><nowiki>
 +
qemu/x86_64-softmmu/qemu-system-x86_64 -boot c -hda /images/xpbase.qcow2 -m 384 -net nic,model=virtio -net tap,script=/etc/kvm/qemu-ifup
 +
</nowiki></pre>
 
* At the moment the kernel modules are automatically loaded in the guest but the interface should be started manually (dhclient/ifconfig)
 
* At the moment the kernel modules are automatically loaded in the guest but the interface should be started manually (dhclient/ifconfig)
 +
* Currently performance is much better when using a host kernel configured with CONFIG_HIGH_RES_TIMERS. Another option is use HPET/RTC and -clock= qemu option.
 +
* Expected performance
 +
** Performance varies from host to host, kernel to kernel
 +
** On my laptop I measured 1.1Gbps rx throughput using 2.6.23, 850Mbps tx.
 +
** Ping latency is 300-500 usec
 +
* Enjoy, more to come :)
  
 
__NOTOC__
 
__NOTOC__

Revision as of 23:27, 31 January 2008

Paravirtualized drivers for kvm/Linux

  • Virtio was chosen to be the main platform for IO virtualization in KVM
  • The idea behind it is to have a common framework for hypervisors for IO virtualization
  • More information (although not uptodate) can be found in kvm AttachFile&do=get&target=kvm_pv_drv.pdf pv driver
  • At the moment network/block/balloon devices are suported for kvm
  • The host implementation is in userspace - qemu, so no driver is needed in the host.

How to use Virtio

  • Get kvm version >= 60
  • Get Linux kernel with virtio drivers for the guest
    • Either build it around Rusty's tree repo
    • Or git clone git://kvm.qumranet.com/home/dor/src/linux-2.6-nv use branch rusty
    • Soon an official repository will be released
    • As an alternative one can use a standard guest kernel for the guest > 2.6.18 and make use sync backward compatibility option
    • Backport and instructions can be found in Anthony Liguori's virtio-ext-modules
    • At the moment it's broken since the guest got developed, soon update
  • Use model=virtio for the network devices.
    • Example
qemu/x86_64-softmmu/qemu-system-x86_64 -boot c -hda /images/xpbase.qcow2 -m 384 -net nic,model=virtio -net tap,script=/etc/kvm/qemu-ifup
  • At the moment the kernel modules are automatically loaded in the guest but the interface should be started manually (dhclient/ifconfig)
  • Currently performance is much better when using a host kernel configured with CONFIG_HIGH_RES_TIMERS. Another option is use HPET/RTC and -clock= qemu option.
  • Expected performance
    • Performance varies from host to host, kernel to kernel
    • On my laptop I measured 1.1Gbps rx throughput using 2.6.23, 850Mbps tx.
    • Ping latency is 300-500 usec
  • Enjoy, more to come :)