Converting OVA for use with KVM / QCOW2
The OVA file is nothing more than a TAR archive, containing the .OVF and .VMDK files. Easy!
Using Evergreen ILS for example:
$ file Evergreen_trunk_Squeeze.ova Evergreen_trunk_Squeeze.ova: POSIX tar archive (GNU)
I'ts possible to use the tar command to list the contents
$ tar -tf Evergreen_trunk_Squeeze.ova Evergreen_trunk_Squeeze.ovf Evergreen_trunk_Squeeze-disk1.vmdk
Simply extract those things...
$ tar -xvf Evergreen_trunk_Squeeze.ova Evergreen_trunk_Squeeze.ovf Evergreen_trunk_Squeeze-disk1.vmdk
Now take a look at the created files The OVF XML file describes the image, it makes for some interesting reading about the expectations of the running environment.
$ file Evergreen_trunk_Squeeze* Evergreen_trunk_Squeeze-disk1.vmdk: VMware4 disk image Evergreen_trunk_Squeeze.ova: POSIX tar archive (GNU) Evergreen_trunk_Squeeze.ovf: XML document text
Recent versions of qemu are able to run directly from the VMDK file, buy why do that? Use QCOW2, it's better. Execute: qemu-img -h and the last line of output shows the supported formats.
$ qemu-img -h |tail -n1
Supported formats: raw cow qcow vdi vmdk cloop dmg bochs vpc vvfat qcow2 parallels nbd blkdebug sheepdog host_cdrom host_floppy host_device file
Now actually convert it, this may take some time.
$ qemu-img convert -O qcow2 Evergreen_trunk_Squeeze-disk1.vmdk Evergreen_trunk_Squeeze.qcow2
Contents of OVF Files
The OVA is intended to run an Appliance and this OVF file describes the appliance. Examine the contents of this file to determine information about the expected CPU, Memory and other appliance details. These will be important to getting the image to run under KVM. Noteably, Windows has a terrible time moving, ensure you have the MergeIDE fix in place.