P2V large FileServer

[egroeg1]

I have to P2V a large file server on one of my next projects - roughly 4TB of file shares on the D drive.

Typically I create two LUNs on my SAN, Once for the C Root and one for D Data, then run converter to transfer the physical disks into those two seperate LUNs as VMFS VMDKs.

As this will be a large file store, I am intending on having my C as a VMFS LUN and creating an iSCSI link to my Data LUN which I will keep as NTFS mapping from the Windows VM.

1. My question is, what is my best route of attack.

2. At present, my current intention is to P2V the windows C drive,

3. Boot the VM, create the iSCSI mapping to the empty D Lun

4. Ghost the D partiton from the old physical server to the new D Lun

5. Restart Windows shares in the VM which will connect to the correct disk

Or is there a more suitable and recommended method that I am missing?

Regards

[AWo]

For C: I would use the Converter. For the data partition I would prepare the virtual disk or RDM and attach is to a different virtual machine temporarily (as you can't have the virtual clone online at the same time as the physcial system) and use "robocopy" to get the data across over time and at the last step cut the users off, do a delta copy and switch the disk to the virtual clone and present that one to the network.

AWo

VCP 3 & 4

\[:o]===\[o:]

=Would you like to have this posting as a ringtone on your cell phone?=

=Send "Posting" to 911 for only $999999,99!=

[jaydub45]

I agree with AWo. I have migrated several large servers this way.

I actually create a new VM from my Windows 2003 or 2008 Template, and then use a tool called "Secure Copy" to synchronise the data. I run it during the day over a period of a few days. I run it once and it copies all the data (it copies the NTFS permissions, takes ownership if it has to and then sets the permissions back on the destination, and also re-creates the shares on the destination), and then run it again, and again over the next few days and configure it to do a synchronise. Once my outage window rolls around, I run a final sync (on a server with 3TB of data this final sync took about 3 hours as it has to scan and compare every single file, but original sync I did during the day took over 23 hours to complete, I can't imagine doing that in my change window).

Once the sync is complete, I shut down the original, rename the OS in the VM and I'm good to go. The downtime for the users is a few hours instead of a day or more.