Skip to main content
Hi,

I have a standalone ESX 5.1 server with a VM with multiple 2TB (screenshot from WinScp attached below) vmdk volumes.

VM is Windows 2008 R2.



My Nutanix cluster is running 5.01, NCC 3.0.0.2, AHV 20160925.30 - Starter Edition



I am planning to use a Windows 2012 R2 server with WinSCP to get the vmdk files onto the Nutanix cluster container (I have multiple containers) and then use the Image service (via Chrome browser) to upload/convert via UNC to the container.







Here are my questions:


  1. I'm wondering if there are any limitations on the Image Server when using the Chrome Browser
  2. Any potential issues with these vmdk sizes?
  3. What is the syntax for the URL to access the Storage Container so I can upload the vmdk without using the UNC? -Figure this one out. In Image Service use nfs://clusterip/Containername/name ofvmdk-flat.vmdk
  4. Since I need to upload multiple vmdk files that quite large, can/should I open multiple browsers to get simultaneous uploads happening ?


TIA,



Genito







If you just SCP them on to the cluster, you shouldn't have to run them through the image service, as the -flat.vmdk format is actually the same between the two systems, so you can just "clone from ADSF image" when creating the disk, and use the -flat as your source.



That will save you heaps of conversion time, and would bypass all of your questions here.
Jon,



Thanks so much!!!

I just did another server with a 1.8TB vmdk and it took a day+ to upload and then image create!!!!



Serves me right!!!!, I should have asked before and could have saved myself a weekend....



Lesson learned....



G
I'd encourage you to test it on a small VM, perhaps a test/template, something tiny, just to get the process down. You'll see its pretty straight forward, but worth just testing out to make sure you know 100%.



Jon
Jon,



Thank I will test it with smaller vmdk.



Does this work with HyperV vhdx as well ?



Thanks,

G
No, everything else needs to go through image service
I have a 12TB OVA file untarred into 2TB vmdks:



ISE-2.2.0.470-eval-disk1.vmdk.gz.000000000

ISE-2.2.0.470-eval-disk1.vmdk.gz.000000001

ISE-2.2.0.470-eval-disk1.vmdk.gz.000000002

ISE-2.2.0.470-eval-disk1.vmdk.gz.000000003

ISE-2.2.0.470-eval-disk1.vmdk.gz.000000004

ISE-2.2.0.470-eval-disk1.vmdk.gz.000000005

ISE-2.2.0.470-eval.mf

ISE-2.2.0.470-eval.ova

ISE-2.2.0.470-eval.ovf



How do you get this running on Nutanix?
Hi Jon!



I'm currently trying to import / move a vmdk from vmware esxi6 to our nutanix datastore;

- I can nfs-mount the nutanix datastore within vsphere client flawlessly...

- I can copy the folder containing vmdk vmx files etc from vmware

BUT

- If I try to clone the correctly copied vmdk in prism via "Clone ADSF image" I have to enter disk size (?)

- If I enter the same disk size, it doesnt boot correctly; the used size of the virtualdisk (256GB) is just some byte

- Which input formats does the image service accept, i mean

... I don't have any vmdk-flat file just the regular 256GB vmdk from the esxi host, does it need the whole foldercontent of the vm from esxi any ideas how to import the vmdk?
Hi mcdean

i think you have to concatenate the split vmdk images first

eg.

https://vmexpo.wordpress.com/2014/04/15/how-to-merge-multiple-vmdks-into-single-vmdk/



I personally wouldn't do this on a windowsmachine, given the fact that multiple 2tb files would need some kind of ReFS setup first (I guess) ..



You basically open up vsphere client, mount the nutanix datastore via nfs, copy over the vm, ssh into ntnx and concatenate with cp first but I'm just a forumuser and not sure if Jon can acknowledge any of my recommendations ...