Jacob,

If you want to use the vdi image (as opposed to a zvol), you'll need
to 'vdiskadm import' it.  This will place the vdi into the format
that xvm understands (basically it creates a directory containing
the image and generates an xml file).

# vdiskadm help import

import - import a zvol, raw file, or virtual disk

USAGE:
 vdiskadm import [-fnpqm] [-x <type>] -d <file|zvol|dsk> [-t <type[:opt]>] 
vdname

EXAMPLE:
  vdiskadm import -d /downloads/image.vmdk /export/new_guests/disk1


This example will copy the vdi file to the /export/new_guests/disk1 directory.
If you don't have room for a copy, you can use
'vdiskadm import -m -d /downloads/image.vdi /export/new_guests/disk1'
(where m == move).

You'd then specify:
disk = ['tap:vdisk:/export/new_guests/disk1,xvdc,w']

Susan




Jacob Ritorto wrote:
Apparently Xen 3.1.4 dislikes this VirtualBox image for some reason.  Here's my 
process and result:

1) I prepared the comms vdi according to the instructions at the download site. 
 First few blocks look like this when viewed through less(1).

<<< Sun xVM VirtualBox Disk Image >>>
^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^...@^@^?^pھ...@^a^@<90>^...@^@^A^@

2) for simplicity, I copied it to a local file called Hawkeye.vdi

3) I created a Xen config file and used xm new to pull it in (don't hate me -- 
I had to do it this way to make NetWare work and I haven't learned how to do 
the new stuff yet):
j...@adra~[116]11:13>cat Hawkeye.xmconfig vcpus = 2
memory = "2048"
name = "Hawkeye"
vif = [ '' ]
vfb=[ "type=sdl" ]
disk = ['tap:vdisk:/jack/Hawkeye.vdi,xvdc,w']

4) ran xm start Hawkeye and watched it puke on the image:
j...@adra~[111]11:05>pfexec xm start Hawkeye
Error: Boot loader didn't return any data!
Usage: xm start <DomainName>

Start a Xend managed domain
-p, --paused Do not unpause domain after starting it -c, --console_autoconnect Connect to the console after the domain is created


So..  Any insight?

_______________________________________________
xen-discuss mailing list
[email protected]

Reply via email to