Hello, I'm trying to export a VM as an OVA from an ESXi 4.0 host through our vCenter Server again version 4.0. At present my command looks like this:

ovftool.exe vi://<IP Address>/< Datacenter>/vm/<Folder Name>/<Folder Name>/<Folder Name> C:/<Path To OVA Store>/<VM Name.OVA>

The final folder name given above is the location of the VMX / VMDK files for the VM. I'm using version 2.1.0-467744-win-i386 of the ovftool from a Windows XP VM running SP3.

Issuing the above command appears to kick off the ova export process but it quickly dies and returns my prompt. The only output after issuing the command is a request for a username and password which I supply followed by the words "Opening VI source".

Any ideas where I'm going wrong?

Hi all,

Is there a programmatical way to exchange data (key-value pairs) with the guest os? i.e. for windows guests, I want to query certain section of the guest os registry. Somewhere i remember reading that this is possible using OVF tools. Any ideas?

Hi All,

Everytime I manually export a VM using vsphere client, I am getting the resultant ovf of much larger size. ( i.e. size of .vmdk's are much larger, increases by 10% everytime )
I looked at various temporary files and log files and deleted them before taking the ovf, but it also doesnt reduces the size significantly.
My vm doesnt do any huge IO on the disks.

Is there a best practice for exporting an ovf .? ( Ex. SLES based vms. )

Thanks

Nikunj

Hi,

     I try to upload a physical server image to vcloud but it keep returning this error "Invalid response from Server"

The ovftool version is 3.01 and the vcloud is 1.5.

I have also attahed the ovf file.

Thank you.
Regards,

ET Goh

hi everyone

im trying to install Solar Winds virtualization manager on my Esxi 3.5.

The appliance is on OVA format.

i got o File--> Virtual Appliance -->Import-->Import the file and i get the next error:  An error occurred while parsing the OVF descriptor: System out of memory exeption.

Then i try to extract the ova file, and i get 4 new files, one of them a OVF. then i try to use it and i get the the next error: An error occurred while parsing the OVF descriptor : Unable to parse xml element ´Envelope´.

This is a OVF version error, 3.5 use a older version of OVF, (0.9 i think).

I dont know how to install this appliance , i hope u can help me.

SORRY FOR MY BAD ENGLISH!!!

Hello There,

OvfTool: 3.0.1

vCloud Director(VCD): 5.1

I use ovftool 3.0.1 to upload an ovf file to VCD and the server return error after upload completely.

It works previously while use ovftool 2.1 with VCD 1.5

I use ovftool 3.0.1 command line to upload template like this:

ovftool --X:logFile=upload.log --machineOutput --powerOn --acceptAllEulas --ipAllocationPolicy=transientPolicy "--net:mynetwork01=mynetwork01" --name=test88t "/tmp/ova/vwin1/vwin1.ovf" "vcloud://xxx:xxx@mynetwork/?org=zhongnan&vappTemplate=test88t&vdc=zhongnan-ALLOC-02&catalog=ISO%20Library" | while read x ; do echo $x; done

The ovf uploaded to VCD is complete, however I got an error result:

<?xml version="1.0" encoding="UTF-8"?>

<InstantiateVAppTemplateParams name="test88t" deploy="false" powerOn="false" xmlns="http://www.vmware.com/vcloud/v1.5"

                               xmlns:ovf="http://schemas.dmtf.org/ovf/envelope/1">

  <InstantiationParams>

    <NetworkConfigSection>

      <ovf:Info>Configuration parameters for logical networks</ovf:Info>

      <NetworkConfig networkName="mynetwork">

        <Configuration>

          <ParentNetwork href="https://mynetwork/api/network/xxx-xxx-xxx-xxx"

                         type="application/vnd.vmware.vcloud.network+xml" name="mynetwork-01"/>

          <FenceMode>bridged</FenceMode>

          <AddressingMode>POOL</AddressingMode>

        </Configuration>

        <IsDeployed>true</IsDeployed>

      </NetworkConfig>

    </NetworkConfigSection>

  </InstantiationParams>

  <Source href="https://mynetwork/api/vAppTemplate/vappTemplate-aaa-bbb-ccc-ddd"/>

  <AllEULAsAccepted>true</AllEULAsAccepted>

</InstantiateVAppTemplateParams>

        2013-06-11T04:33:17.786-07:00 [7FA9BA208720 verbose 'Default'] VCloud response body:

        --> <?xml version="1.0" encoding="UTF-8"?>

        --> <Error xmlns="http://www.vmware.com/vcloud/v1.5" minorErrorCode="BAD_REQUEST" message="Bad request

-->  - Unexpected JAXB Exception

-->  - cvc-complex-type.2.4.a: Invalid content was found starting with element 'AddressingMode'. One of '{&quot;http://www.vmware.com/vcloud/v1.5":RetainNetInfoAcrossDeployments,

&quot;http://www.vmware.com/vcloud/v1.5":Features, &quot;http://www.vmware.com/vcloud/v1.5":d, &quot;http://www.vmware.com/vcloud/v1.5":RouterInfo}' is expected."

                   majorErrorCode="400" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

                   xsi:schemaLocation="http://www.vmware.com/vcloud/v1.5http://sd1.aiscloud.net/api/v1.5/schema/master.xsd"></Error>

        -->

        2013-06-11T04:33:17.788-07:00 [7FA9BA208720 verbose 'Default'] Backtrace:

        --> backtrace[00] rip 00007fa9b42a0f44 Vmacore::System::Stacktrace::CaptureWork(unsigned int)

        --> backtrace[01] rip 00007fa9b406bca4 Vmacore::System::SystemFactoryImpl::CreateQuickBacktrace(Vmacore::Ref<Vmacore::System::Backtrace>&)

        --> backtrace[02] rip 00007fa9b4001ef8 Vmacore::Throwable::Throwable(std::string const&)

        --> backtrace[03] rip 000000000061e7c1 /usr/lib/vmware-ovftool/ovftool.bin() [0x61e7c1]

        --> backtrace[04] rip 000000000060fca8 /usr/lib/vmware-ovftool/ovftool.bin() [0x60fca8]

        --> backtrace[05] rip 000000000064eb65 /usr/lib/vmware-ovftool/ovftool.bin() [0x64eb65]

        --> backtrace[06] rip 000000000064fd7a /usr/lib/vmware-ovftool/ovftool.bin() [0x64fd7a]

        --> backtrace[07] rip 0000000000569c84 /usr/lib/vmware-ovftool/ovftool.bin() [0x569c84]

        --> backtrace[08] rip 00000000004c4b6e /usr/lib/vmware-ovftool/ovftool.bin() [0x4c4b6e]

        --> backtrace[09] rip 00000000004c7be4 /usr/lib/vmware-ovftool/ovftool.bin() [0x4c7be4]

        --> backtrace[10] rip 00007fa9b1cb1c4d /lib/libc.so.6(__libc_start_main+0xfd) [0x7fa9b1cb1c4d]

        --> backtrace[11] rip 00000000004a2f79 std::basic_streambuf<char, std::char_traits<char>>::xsgetn(char*, long)

Please let me if this is a bug in ovftool to integrate VCD 5.1.

Thanks,

Jeff

I have a setup where I am accessing vCenter and an ESXi host through a NAT layer. If I try to deploy directly to vCenter, ovftool automatically tries to connect to the ip address of the ESXi host behind the NAT layer, which fails.

It looks something like this:

[root@ovfimporter mnt]# date; ovftool --datastore="PayGo Storage" --net:"VLan 4=DedicatedPortGroup_NNNN" CUST/CUST.ovf vi://matthew.wallace@10.10.10.1/Customer\ Datacenter/host/PayGo\ Cluster/Resources/Paygo/NNNN\ \(abc44060-35dc-438a-8c79-6bc7586082ab\)/ ;date

Tue Oct  8 10:18:43 MDT 2013

Opening OVF source: CUST/CUST.ovf

The manifest validates

Enter login information for target vi://10.10.10.1/

Username: matthew.wallace

Password: *********

Opening VI target: vi://matthew.wallace@10.10.10.1:443/Customer%20Datacenter/host/PayGo%20Cluster/Resources/Paygo/NNNN%20(abc44060-35dc-438a-8c79-6bc7586082ab)/

Deploying to VI: vi://matthew.wallace@10.10.10.1:443/Customer%20Datacenter/host/PayGo%20Cluster/Resources/Paygo/NNNN%20(abc44060-35dc-438a-8c79-6bc7586082ab)/

Transfer Failed

Transfer Failed

Error: Lost connection to target

Completed with errors

Tue Oct  8 10:20:19 MDT 2013

Looking at netstat, I saw a hung connection to 443 of one of the ESX hosts in that PayGo Cluster:

In this case, 172.17.0.10 is the "real" IP of one of the esx hosts in the PayGo cluster managed by that vCenter, which I can reach from the client running the ovftool only by using the NAT ip of 10.10.10.100.

So, I tried to specify the external ip of that host. Per the ovftool 3.5 user guide:

Given that, I tried using several syntaxes:

ovftool --datastore="PayGo Storage" --net:"VLan 4=DedicatedPortGroup_NNNN" CUST/CUST.ovf vi://matthew.wallace@10.10.10.1/Customer\ Datacenter/host/PayGo\ Cluster/Resources/Paygo/NNNN\ \(abc44060-35dc-438a-8c79-6bc7586082ab\)/?ip=10.10.10.100

or even just the simplest version to see if I can get a prompt for going further down in the locator tree:

ovftool --datastore="PayGo Storage" --net:"VLan 4=DedicatedPortGroup_NNNN" "vi://matthew.wallace@10.10.10.1/?ip=10.10.10.100"

I invariably get a result like this:

The manifest validates

Enter login information for target vi://10.10.10.1/

Username: matthew.wallace

Password: *********

Error: Locator does not refer to an object: vi://matthew.wallace@10.10.10.1:443/?ip=10.10.10.100

Completed with errors

Outside of the brief mention of the "ip=xxx" param mentioned in the ovftool userguide I've been unable to find any reference to someone successfully using this parameter so I'm uncertain if this is a bug, my not understanding the intention of the parameter as specified in the OVF tool guide, or me using bad syntax.

I am facing an issue. I have an ova file of size 362 MB. I am using OVFTOOL to convert ova to vmx file. I get the below error during conversion,

Also tried with option --chunkSize=200mb, but with no luck. Same error as below.

Transfer Failed

Error: System error: (Unknown exception)

Completed with errors.

In Vmware workstation 8, i get the below error when i am trying to import the OVA file with the below steps,

a). From the File menu on the VMware Workstation 8 menu bar, select Open.

b).  Browse to the .ovf or .ova file and click Open and Import.

I get the below error.  Anyone who could help me with this?

Hi,

trying to convert a "SUSEstudio" generated vmdk/vmx to our VI.

I have:

SLE_11.1JeOSPHR.x86_64-0.0.3.vmdk

SLE_11.1JeOSPHR.x86_64-0.0.3.vmx

christian@chrisw:/srv/ISO/JeOS/SLE_11.1JeOSPHR-0.0.3&gt; ovftool SLE_11.1JeOSPHR.x86_64-0.0.3.vmx

Opening VMX source: SLE_11.1JeOSPHR.x86_64-0.0.3.vmx

OVF version:   1.0

Name:          SLE_11.1JeOSPHR.x86_64-0.0.3

Download Size:     Unknown

Deployment Sizes:

  Flat disks:      4.04 GB

  Sparse disks:   Unknown

Networks:

  Name:        bridged

  Description: The bridged network

Virtual Hardware:

  Family:       vmx-04

  Disk Types:   SCSI-lsilogic

Completed successfully

#####

When I try to convert it to vi (to correct datastore) I'm getting the mentioned error:

so what is the correct syntax ??

christian@chrisw:~/srv/ISO/JeOS/SLE_11.1JeOSPHR-0.0.3&gt; ovftool -ds=CX300-OTC-02 -dm=thick SLE_11.1JeOSPHR.x86_64-0.0.3.vmx vi://i000156@vc.icw.int/DC%20OTC

Opening VMX source: SLE_11.1JeOSPHR.x86_64-0.0.3.vmx

Please enter login information for target vi://vc.icw.int/

Username: i000156

Password: ********

Error: Locator does not refer to an object: vi://i000156@vc.icw.int/DC%20OTC

DataCenter="DC OTC"

DataSource="CX300-OTC-2"

THX

Chris

Hi,

I am trying to deploy OVA using ovftool. I see the option --network and --net available for selecting the network by name, though I was not able to find a way to assign a static IP in the ovftool command itself. Something like --ip0=10.10.10.10. Is there a way we can do this? From UI when we deploy an OVA, we have provision to provide IP, Gateway, Netmask and DNS.

Thanks

Hello all,

I am having a problem with the ovftool.

Here's what i did.

1)I created a VM running CentOS.

2)These are the 3 files that i got.

Test.vmx

Test.vmdk (Descriptor File)

# Disk DescriptorFile

version=1

CID=5046e880

parentCID=ffffffff

createType="vmfsPreallocated"

# Extent description

RW 83886080 VMFS "Test-flat.vmdk" 0

# The Disk Data Base

#DDB

ddb.virtualHWVersion = "4"

ddb.geometry.cylinders = "83220"

ddb.geometry.heads = "16"

ddb.geometry.sectors = "63"

ddb.adapterType = "lsilogic"

Test-flat.vmdk

3) This is the command that i ran

ovftool "Test.vmx" "Test_New.ova"

Please help!

Thank you

Running OVF Tool 3.0.1 on 32 bit CentOS 6 Linux 2.6.32-279.11.1.el6.i686 (download from VMWare site md5sum was checked ok)

Tool starts up fine and is in the process of exporting a VM from one ESXi 4.1 host to another ESXi 4.1 host . 

Tool then crashes with /usr/bin/ovftool: line 23:  7629 Segmentation fault      (core dumped) "$OVFTOOL_BIN" "$@"

Hi!

I'm trying to to deploy a OVF/OVA file into my vCenter but I'm getting a error message, pls help me.

"Line 61: No space left for device '5' on parent controller '4'."

(Tested to deploy other templates and it worked fine)

Any idea anyone?

Using ovftool 3.0.1, I cannot figure out how to escape the backslash character when specifying the username and password in an FTP locator on the command line.

$ ovftool ftp://<DOMAIN>%5C<username>:<password>@<host>/Image.ova

keeps prompting me for the password, even if I repeatedly enter the correct one.

The same thing happens when I specify "ftp://<DOMAIN>%5C%5C<username>:<password>@<host>/Image.ova" instead.

Using the literal backslash character (i.e. <DOMAIN>\<username> or <DOMAIN>\\<username>) results in an error message saying "Error: cURL error: Couldn't resolve host name."

However, if I do not provide any credentials for the locator, ovftool prompts for the username and password. In that case, providing "<DOMAIN>\<username>" does work, though "<DOMAIN>%5C<username>" does not.

Please note that the FTP server unfortunately will not properly authenticate without prepending the domain.

Installed ovftool 3.5 on CentOS 5.9.  Install went without any errors, says Installation Successful.

Running ovftool yields about 5 errors similar to:

/usr/lib/vmware-ovftool/ovftool.bin: /usr/lib64/libstdc++.so.6: version `GLIBCXX_3.4.9' not found (required by /usr/lib/vmware-ovftool/libvmomi.so)

CentOS 5.9 doesn't support GLIBC 3.4.

The ovftool documentation says it supports CentOS 5 and 6.

What's the problem here?

Hi VMWare experts,

I am having problem with automation script because the string passed in the ovftool command uses credential vsphere.local\userXYZ instead of just userXYZ

It appears that the backsplash in the 'vsphere.local\user' cause the command line to be misinterpreted, so the automation script that calls ovftool to deploy VMs kept failing.  It erred with "cURL error of ‘couldn’t resolve hostname" even though there is nothing wrong with any hosts or hostname resolution on the network.

On a Win-based vCenter I was able to reconfigure it so that instead of requiring login user format of 'vsphere.local\userXYZ', it will now accept user = 'userXYZ'. 

With that change, ovftool.exe command string is now progress correctly and VMs are deployed. 

I could reconfigure Window-based vCenter to not use just user = userXYZ (instead of vsphere.local\account), but I could not figure out how to get vCSA to stop requiring the format of vsphere.local\userXYZ.

The string segment in the long ovftool.exe command line looked like this …

This works:

"vi://userXYZ:password@tm-vcenter55/DC2-SJ/host/Topology/csg-sj-esxi112.cisco.com"

This does not work:

"vi://vsphere.local\userXYZ:password@lin-vcenter55/DC2-SJ/host/Topology/csg-sj-esxi111.cisco.com"

Is there anyway to configure vCenter Virtual Appliance to use log in format of user = userXYZ instead of user = vsphere.local\userXYZ?   Or is there some way of reformatting the string so that OS can interpret it correctly even with backsplash '\' is used in the string?

Thank you

Charles

The complete ovftool.exe command is shown below:

"C:\Program Files\VMware\VMware OVF Tool\ovftool.exe" --acceptAllEulas --skipManifestCheck --diskMode=thin --name=linsys2-vm1 --deploymentOption=vm_typeA --datastore=EMC-LUN1 --net:"Network Adapter 1"="VLAN 410 - 172.27.14.0/24" --net:"Network Adapter 2"="VLAN 410 - 172.27.14.0/24" --vmFolder=   --prop:cisco_hostname.Proj=linsys2-vm1 --prop:cisco_ip_0.Proj=172.27.14.135 --prop:cisco_dnsDomain.Proj=cisco.com --prop:cisco_netmask_0.Proj=255.255.255.0 --prop:cisco_gateway_0.Proj=172.27.14.1 --prop:cisco_dns1_0.Proj=172.27.200.199 --prop:cisco_dns2_0.Proj=173.36.131.10 --prop:cisco_language.Proj=English v:\ova\projectX_1.0.0.ova "vi://vsphere.local\userXYZ:password@lin-vcenter55/DC2-SJ/host/Topology/csg-sj-esxi111.cisco.com"

Opening OVA source: v:\ova\projectX_1.0.0.ova

The manifest does not validate

Error: cURL error: Couldn't resolve host name

Completed with errors

Installation Done!

Hi all,

I am trying to implement an OVF support for some virtualization software, which is using raw images for disks. I use qemu-img to convert raw images to vmdk format. Then I use ovftool to convert ovf file into vmware instance.

When the file I have is a simple ovf descriptor + manifest + disk - all is fine, it works even if I'm using qemu-img without "compat6" option

When the disk is compressed or the OVF package is packed into OVA - there are errors being thrown. Moreover when I create OVA package with no disk, the conversion works.

Here is the fragment of results for the OVA file:

Writing VMX file: test_ova.vmx
[2011-10-03 16:34:35.685 7F097D2C1720 verbose 'ovftool'] Using file size as progress measure
Disk progress: 0%[2011-10-03 16:34:35.685 7F097D2C1720 verbose 'ovftool'] Ovf::GetFileSource as StreamVmdk failed. Exception: Cannot open disk source: Not a supported disk format (sparse VMDK version too old)
Disk Transfer Failed                      
Error: Failed to open source disk: 3f66f54d-a9fa-4aa8-806d-d757d6f34a59.disk0.vmdk
[2011-10-03 16:34:35.685 7F097D2C1720 verbose 'ovftool'] Backtrace:
[00] rip 00007f09767453e4 
[01] rip 00007f09765e99d4 
[02] rip 00007f097658e038 
[03] rip 00000000004fef51 
[04] rip 000000000054aaa5 
[05] rip 000000000054ae78 
[06] rip 000000000049c41a 
[07] rip 00007f0974e62c4d 
[08] rip 000000000047ec7a 

[2011-10-03 16:34:35.685 7F097D2C1720 verbose 'ovftool'] Deleting file: ./test_ova-disk1.vmdk

Same fragment when I unpack OVA and then run ovftool on the .ovf file:

Writing VMX file: test_ovf.vmx
[2011-10-03 16:35:53.934 7F7E899C4720 verbose 'ovftool'] Using file size as progress measure
Disk progress: 0%[2011-10-03 16:35:53.935 7F7E899C4720 verbose 'ovftool'] Ovf::GetFileSource as StreamVmdk failed. Exception: Cannot open disk source: Not a supported disk format (sparse VMDK version too old)
Disk progress: 99%[2011-10-03 16:35:54.106 7F7E899C4720 verbose 'ovftool'] Ovf::GetFileSource: Opened 3f66f54d-a9fa-4aa8-806d-d757d6f34a59.disk0.vmdk as diskLib source
[2011-10-03 16:35:54.106 7F7E899C4720 verbose 'ovftool'] Opening disk target /xcinstall-1.corp.google.com/VirtualLsiLogicController0:0
Disk Transfer Completed                   
Completed successfully

Thank you.

Hi,

I'm getting the following error when trying to convert from VMX to OVF:

1. ovftool ./n1dps/Solaris\ 10\ 64-bit\ Single\ Partition.vmx ./ovf/n1dps.knerr.us.ovf

Opening VMX source: ./n1dps/Solaris 10 64-bit Single Partition.vmx

Error:

- Device 'cdrom1' has a backing type that is not supported. This is a general limitation of the virtual machine's virtual hardware version on the selected host.

System Information:

Host: CentOS 5.0 64bit

Vmware Server version=1.0.6 build=build-91891 option=Release.3.2.2

Instance: Solaris 10

I notice the target is Vmware Server 2.x and upgrading is on the task list.  I thought I'd post this in the mean time and see if anyone had any answers.

Thank you,

George

Update: I just removed the cdrom1 device and am now in the process of running the script as it did not give the cdrom1 error this time.

I am trying to import OVF with custom propery set and getting folling error with vmware ovf tool. Any idea on how to resolve this issue?

Error:

- Line 107: Unsupported element 'Property'

Folks,

We will be moving this community into the Automation Tools Community within the VMTN.

New Location: http://vmware.com/go/sysadmintools

When: Thursday May 27th 12 Noon PST

We expect impact to be low. Please let us know if you have any links you need to preserved.

Thanks