Hi Stan
>>> We have a set of images sent to us by a customer.
I noticed that you processed those images with different tools - probably this is a completely wasted effort.
If your customer has created the images with dd and if he provides disk images rather than partition images you dont need to modify the images at all.
In fact a diskimage created with dd is already in an ESXi-compatible format. Just rename it as image-flat.vmdk.
You can create an appropriate descriptorfile easily: just use this sample and edit 3 variables
### Replace
### $X with the size of the dd image in sectors (size in bytes / 512)
### $Y with result of $X / 16065 rounded down
### $Z with the name of the dd-image
### Keep everything else as it is
# Disk DescriptorFile
version=1
encoding="UTF-8"
CID=12345678
parentCID=ffffffff
createType="vmfs"
# Extent description
RW $X VMFS "$Z"
# The Disk Data Base
#DDB
ddb.adapterType = "lsilogic"
ddb.geometry.sectors = "63"
ddb.geometry.heads = "255"
ddb.geometry.cylinders = "$Y"
ddb.virtualHWVersion = "8"
To verify that the images are dd diskimages and esxi-compatible you can use Linux.
sgdisk -d unknown-imagefile
should output the partitiontable of the disk. If this looks reasonable and matches the overall size of the disk you can assume it is ok to use it without any conversions.
Especially for Solaris none of the known p2v tools is helpful.
For Solaris you should prepare the Solaris VM and use a Solaris LiveCD for the first boot with the dd-image.
The manual modifications that will be necessary require in depth Solaris-skills - or good google-skills.
If you want to make sure you get usable images tell your customers that you want them to prepare the images with a Linux LiveCD.
The Solaris machine should be powered off - booted in to the Linux LiveCD and the image should be created like this for example:
dd if=/dev/sda bs=1M of=hostname-export.dd
If network to an esxi host is available the output can be directly stored in a VMFS-volume if you run this commands first
mkdir /vmfs-out
sshfs root@esxi-host:/vmfs/volumes/datastoreX/ /vmfs-out/
mkdir /vmfs-out/solaris-import
dd if=/dev/sda bs=1M of=/vmfs-out/solaris-import/hostname-export-flat.vmdk
This works very well and is very predictable.
Actually I use this with some modifications to do Coldclone P2V actions that only allow a very short downtime before the transition from P to V has to go active.
Ulli