Unable to Import Wilds from vCenter

Hello Everyone,

I have installed OpenNebula Community edition 6.0.0.2 on CentOS 7. I followed the installation guide given by OpenNebula & everything goes well on the installation.

I have vCenter 6.5 and EXS 6 nodes. I have separate SAN environments for storage and current running VMs are stored on them.

I imported all Hosts, Clusters, Networks & Datastores from vCenter to the OpenNebula FrontEnd server without any error.

But when I’m going to import Wild VMs from existing Hosts, I’m getting “Wild with ref vm-31475 could not be imported. Error creating disk from template: [one.image.allocate] Cannot parse image SIZE”.

Can you please help me to sort this problem?

Please let me know if there need any information for further

Regards
Himash

Currently we do not support ESX with OpenNebula 6.0, you could update your ESXi to the minimum version 6.5 to see if that solves your problem. And please let us know to stay up to date.

1 Like

Hi @cherrera,

Sorry I have missed letter “I” on ESX & we have use ESXI 6 nodes on our POC environment which is very similar setup to planning production OpenNebula environment.

Some VMs (wild) can import through the Sunstone or CLI and most of the VMs can’t. On that point I think there is not any version compatible issue with the Hypervisor & the OpenNebula platform. Please correct me if I’m wrong.

I have attached /var/log/oned.log output for the mean time of trying to import the Wilds.

Fri Sep 3 12:09:13 2021 [Z0][ImM][E]: stat: Error calculating image vcenter://Himash-CentOS-[LAN]/Himash-CentOS-[LAN]_1.vmdk size. Reason: “Could not find file.”.
Fri Sep 3 12:09:13 2021 [Z0][Lis][I]: Stopping
Fri Sep 3 12:09:13 2021 [Z0][ReM][E]: Req:4224 UID:0 one.image.allocate result FAILURE [one.image.allocate] Cannot parse image SIZE:

Best Regards
Himash

We advise to use a compatible version of ESX with OpenNebula 6, the minimum being vSphere 6.5 (both vCenter and ESX). We strongly discourage the use of a previous versions with OpenNebula for a production environment.

Having said that, this error might not be related with the version. We can find more about this particular error enabling debug in /var/lib/one/remotes/etc/vmm/vcenter:

:debug_information: true

let us know what shows in the log files after activating this debug, related with the stat error.

1 Like

@tinova thank you for your prompt reply. Please find the below

Fri Sep 3 16:16:53 2021 [Z0][ImM][E]: stat: Error calculating image vcenter://Himash-CentOS-[LAN]/Himash-CentOS-[LAN]_1.vmdk size. Reason: “Could not find file.”.
Fri Sep 3 16:16:53 2021 [Z0][ImM][I]: Error calculating image vcenter://Himash-CentOS-[LAN]/Himash-CentOS-[LAN]_1.vmdk size. Reason: “Could not find file.”. ["/usr/lib/one/ruby/vcenter_driver/datastore.rb:562:in rescue in stat'", "/usr/lib/one/ruby/vcenter_driver/datastore.rb:543:in stat’", “/var/lib/one/remotes/datastore/vcenter/stat:76:in `’”]
Fri Sep 3 16:16:53 2021 [Z0][Lis][I]: Stopping
Fri Sep 3 16:16:53 2021 [Z0][ReM][E]: Req:9664 UID:0 one.image.allocate result FAILURE [one.image.allocate] Cannot parse image SIZE:

Apologies for not noticing before, currently it is not allowed to have have special characters in the name of the disks as well as in their paths.

Image names and paths cannot contain spaces or non-ASCII characters.

As a solution to import the virtual machine, please rename the disks in vCenter and try again.

1 Like

@cherrera & @tinova Thank you very much for helping me sort this out. Much appreciated :ok_hand:

1 Like