After disabling all that ‘Smartware’ software as explained by on Western Digital’s website, I still had failures installing ESXi 4.1 when mcopy tries to grab the kickstart file.
And after loading all you get to see is:
rescanning in 10 second(s), press
This happened with another drive I had so I finally got down and dirty into it.
WD as well as others put a GPT partition on the first and last sectors of the drive. I found this while looking for something else here. So zeroing out the drive seems to do the trick. Here is what you do:
1. Connect the Drive to a Linux machine and do fdisk
# fdisk -l
Disk /dev/sde: 750.1 GB, 75017153152 bytes
255 heads, 63 sectors/track, 91197 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
2. Zero out some bytes
Take note of the number of cylinders and the units. In my example there were 91197 cylinders broken down by 16065. So first zero out the first couple of bytes on the disk:
# dd if=/dev/zero of=/dev/sde bs=512 count=2
Now do some math: multiply the number of cylinders by the unit size. For me its 91197 * 16065 = 1465079805 (this is intense math people). Now use that:
# dd if=/dev/zero of=/dev/sde bs=512 seek=1465079805
3. format your disk as I explained on my other post
4. Mod voodoo
The next thing is that you’ll have to add a mod.tgz or oem.tgz file that contains an init script that creates a ~/mtoolsrc file to ignore the bad offsets. Here’s an easy one to do:
echo “mtools_skip_check=1” >/.mtoolsrc
Pack that up and add it to your syslinux.cfg at the end. You should be good to go from that point on. I’d be interested to know if this helps anyone. Any feedback, errors, or grips would be great. Thanks.