Difference between revisions of "Migration from VMware"

From Xen
(Extract VM from the VMware Datastore)
(Disk Conversion)
Line 22: Line 22:
 
== Disk Conversion ==
 
== Disk Conversion ==
   
* Halt the VMware VM. Examine the type of file used by the VM:
+
* Examine the type of file used by the VM:
 
<pre>
 
<pre>
 
# qemu-img info vm1.vmdk
 
# qemu-img info vm1.vmdk

Revision as of 01:16, 11 January 2015

As time passes, we are finding more and more people interested in migrating to Xen Project from VMware. The reasons for this migration include cost reduction and increased flexibility in cloud orchestration.

Introduced in the 4.5 release, Xen Project now includes the ability to use VMware VMDK files natively to ease migration. For maximum performance, it is advisable to actually translate the VM files into a raw format. But this step can be delayed, if needed, to facilitate migration.

Extract VM from the VMware Datastore

If the VM in question is located within a VMware datastore, this post suggests a method for making a local copy of the VM to convert:

I would navigate to the datastore where the VM is located.

In the vSphere Client, the datastore is listed under the "Resources" section.

Right-Click the Datastore and choose "Browse Datastore" which should be at the
top of the menu.

Once the Datastore Browser is up, find the VM's folder.

Highlight the folder... Right-Click and choose "Download..."

Disk Conversion

  • Examine the type of file used by the VM:
# qemu-img info vm1.vmdk
image: vm1.vmdk
file format: raw
virtual size: 23G
disk size: 23G
  • In the case above, the file format is raw, so it needs no conversion. Just rename the file to ".raw". However, most of the time, that will not be the case. You will need to use qemu-img to do the conversion.
# qemu-img convert vm1.vmdk -O raw /vms/vm1/vm1.img
  • Create an appropriate Xen Project VM configuration file in /etc/xen/vm1/, with a line specifying the converted file as disk:
disk = [ ‘file:/vms/vm1/vm1.img,hda,w’ ]
  • Now boot and go!