By design I have a seperate Datastore for DMZ. I have copied what I thought was the (.vmx, vdmk) files to this new Datastore. When the files for the VM copied it named them both flat-vdmk. I had read into many posts stating what the flat file is. Question: why did it take the original VMDK and change the name to FLAT.VMDK? I was unable to power up the VM after the copy. If I have this wrong please let me know
1. get the size of the flat file
2. make a new file exact same size using thin provisioning
3. delete the temp file, rename the flat temp (removing the flat portion)
4. then edit the descriptor file, change the .flat file and edit the ddb file.
Seems like a bunch of steps - I tried to rename the flat file to the .vdmk file. (this way I dont have to create another file of the same size on the same disk)
however when I try to point the Hard drives to the .vdmk file the Vsphere doesnt see them in the Datastore directory but they are there and I can see them if I manually Browse the datastore.
1. does VM make the VMDK file into a flat file while copying?
2. Why doesnt Vphere recognize the .vmdk file?
How did you copy the files? And why didn't you just copy the complete folder?
Regarding the virtual disk. Each virtual disk consists of two files, the header file "vmname.vmdk" and the data file "vmname-flat.vmdk" (disregarding snapshot disks in this case) If both files exist, the datastore browser will show them as only a single file with the name of the header file and the size of the data file.
The steps you mentioned (http://kb.vmware.com/kb/1002511) are needed to recreate the header file (which is only a few hundreds bytes in size).
Renaming the flat file does not help at all. If you have trouble recreating the header file, provide the details about the flat file and I will assist you.
I manually exported them using the Datastore tool - Export/Import.
I have decided to do the entire folder in hindsight maybe that would have been a better route.