When I clone a machine to a workstation VM I end up with a block for block copy of the volumes, so if I have a 20 GB volume I end up with a 20 GB vmdk file, even if 10 GB of that volume is free space. Is there any way to limit the size of the resulting vmdk files to the used capacity of each volume? So for instance if I clone a Windows server with a 50 GB C: volume of which 20 GB is free space I would end up with a 30 GB vmdk file for that volume?
Yes, this is possible and to get better results, defrag the machine disks before the conversion.
Here is a blog post showing how to shrink disks using Converter: How to shrink a VMDK using VMware Converter - TechRepublic
Thanks for the reply. The settings I see in my converter don't exactly match what I see in the article. I'm using converter standalone 5.5.
In the article, it has an option for thin provisioned under the type/cluster size, but all I have is preallocated and not preallocated. When I select not preallocated it still creates a vmdk file the same size as the entire volume (including free space). Am I doing something wrong?
Thin provisioning option is available only if destination is VMware Infrastructure, but if you choose the "Not pre-allocated" you will get a dynamic disk, that virtual disk will have the original size but will consume only the amount of data used inside the virtual disk.
I guess my question more clearly stated is why am I getting full size vmdk files if I'm selecting to convert a running machine to a workstation VM and have selected the not preallocated option?
Where you're checking the virtual disk size ? On virtual machine settings or on destination folder ?
On destination folder. I am converting to workstation VMs then transferring via FTP to another location, if I could get thin vmdks it would save a considerable amount of time spent transferring the resulting files.