libvirt users - Aug 2013 - Re: Error while starting the node on ESXi hypervisor

Hi,

Today I was trying to start a node using libvirt on ESX (esxi-5.1)
hypervisor and I got the following error message:

*virsh # define /local/myNode/esxdomain.xml*
*Domain testNode defined from /local/myNode/esxdomain.xml*
*virsh # start testNode*
*error: Failed to start domain testNode*
*error: internal error: Could not start domain: GenericVmConfigFault -
Cannot open the disk
/vmfs/volues/5208f99d-760cf4a2-000c29520788/testNode.vmdk or one of the
snapshot disks it depends on.*
*
*
*
*
I checked the datastore of my ESX server and found out that instead of file
one directory is getting created with the name* cluster.vmdk *and inside it
there are few more files but cluster.vmdk file is not there.
*
*
/vmfs/volumes # cd datastore1
 /vmfs/volues/5208f99d-760cf4a2-000c29520788 # ls
cluster.vmdk
 /vmfs/volues/5208f99d-760cf4a2-000c29520788 # cd cluster.vmdk
 /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk # ls -ltr
*-rw-r--r--   1  root    root   261 Aug 12 17:30 cluster.vmdk*
*-rw-r--r--   1  root    root   674 Aug 12 17:30 cluster.vmx*
*-rw-r--r--   1  root    root   0 Aug 12 17:30 cluster.vmsd*
 /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk #

Can anyone tell me how to start a node and why I am getting that error
message.
Below is my xml file.

*<domain type='vmware'>*
*  <name>testNode</name>*
*  <uuid>50115e16-9bdc-49d7-f171-53c4d7f91710</uuid>*
*  <memory>1048576</memory>*
*  <currentMemory>1048576</currentMemory>*
*  <vcpu>1</vcpu>*
*  <os>*
*    <type arch='x86_64'>hvm</type>*
*  </os>*
*  <devices>*
*    <disk type='file' device='disk'>*
*      <source file='[datastore1] cluster.vmdk'/>*
*      <target dev='sda' bus='scsi'/>*
*      <address type='drive' controller='0' bus='0'
unit='0'/>*
*    </disk>*
*    <controller type='scsi' index='0'/>*
*    <interface type='bridge'>*
*      <mac address='00:50:56:25:48:c7'/>*
*      <source bridge='VM Network'/>*
*    </interface>*
*  </devices>*
*</domain>*

2013/8/12 varun bhatnagar <varun292006@gmail.com>:
> Hi,
>
> Today I was trying to start a node using libvirt on ESX (esxi-5.1)
> hypervisor and I got the following error message:
>
> virsh # define /local/myNode/esxdomain.xml
> Domain testNode defined from /local/myNode/esxdomain.xml
> virsh # start testNode
> error: Failed to start domain testNode
> error: internal error: Could not start domain: GenericVmConfigFault -
Cannot
> open the disk /vmfs/volues/5208f99d-760cf4a2-000c29520788/testNode.vmdk or
> one of the snapshot disks it depends on.
>
>
> I checked the datastore of my ESX server and found out that instead of file
> one directory is getting created with the name cluster.vmdk and inside it
> there are few more files but cluster.vmdk file is not there.
>
> /vmfs/volumes # cd datastore1
>  /vmfs/volues/5208f99d-760cf4a2-000c29520788 # ls
> cluster.vmdk
>  /vmfs/volues/5208f99d-760cf4a2-000c29520788 # cd cluster.vmdk
>  /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk # ls -ltr
> -rw-r--r--   1  root    root   261 Aug 12 17:30 cluster.vmdk
> -rw-r--r--   1  root    root   674 Aug 12 17:30 cluster.vmx
> -rw-r--r--   1  root    root   0 Aug 12 17:30 cluster.vmsd
>  /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk #
Okay, libvirt should not have created a directory named cluster.vmdk,
I think there is a bug, but I can't test it at the moment because my
ESX setup is unavailable due to a hardware failure.
> Can anyone tell me how to start a node and why I am getting that error
> message.
> Below is my xml file.
>
> <domain type='vmware'>
>   <name>testNode</name>
>   <uuid>50115e16-9bdc-49d7-f171-53c4d7f91710</uuid>
>   <memory>1048576</memory>
>   <currentMemory>1048576</currentMemory>
>   <vcpu>1</vcpu>
>   <os>
>     <type arch='x86_64'>hvm</type>
>   </os>
>   <devices>
>     <disk type='file' device='disk'>
>       <source file='[datastore1] cluster.vmdk'/>
Here is a problem. Typically all files of a virtual machine should be
located in a subdirectory named after the virtual machine. But you
placed the VMDK file in the datastore root. I think this confused
libvirt. This should read

<source file='[datastore1] testNode/cluster.vmdk'/>

instead. Also "virsh define" cannot create this VMDK file for you if
it doesn't exist already, because there is not enough information in
the domain XML. You need to do this yourself using "virsh vol-create".

I suggest you delete the cluster.vmdk directory from the datastore and
start over with the domain XML modified as suggested, but create
"[datastore1] testNode/cluster.vmdk" before defining the virtual
machine.

-- 
Matthias Bolte
http://photron.blogspot.com

Thanks a ton for the reply Matthias.
I am little confused here. How to add volume on my esx server using virsh?
I did SCP and copied my vmdk image to [datastore1]  testNode/cluster.vmdk
and also modified my xml file as you suggested. Now when I tried starting
the node it gave me an error message saying:

"error: Failed to start testNode
error: internal error: Could not start domain: GenericVmConfigFault -
Failed to start the virtual machine."

How to resolve this?

Regards,
Varun
 On 13 Aug 2013 13:21, "Matthias Bolte"
<matthias.bolte@googlemail.com>
wrote:
> 2013/8/12 varun bhatnagar <varun292006@gmail.com>:
> > Hi,
> >
> > Today I was trying to start a node using libvirt on ESX (esxi-5.1)
> > hypervisor and I got the following error message:
> >
> > virsh # define /local/myNode/esxdomain.xml
> > Domain testNode defined from /local/myNode/esxdomain.xml
> > virsh # start testNode
> > error: Failed to start domain testNode
> > error: internal error: Could not start domain: GenericVmConfigFault -
> Cannot
> > open the disk
/vmfs/volues/5208f99d-760cf4a2-000c29520788/testNode.vmdk
> or
> > one of the snapshot disks it depends on.
> >
> >
> > I checked the datastore of my ESX server and found out that instead of
> file
> > one directory is getting created with the name cluster.vmdk and inside
it
> > there are few more files but cluster.vmdk file is not there.
> >
> > /vmfs/volumes # cd datastore1
> >  /vmfs/volues/5208f99d-760cf4a2-000c29520788 # ls
> > cluster.vmdk
> >  /vmfs/volues/5208f99d-760cf4a2-000c29520788 # cd cluster.vmdk
> >  /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk # ls -ltr
> > -rw-r--r--   1  root    root   261 Aug 12 17:30 cluster.vmdk
> > -rw-r--r--   1  root    root   674 Aug 12 17:30 cluster.vmx
> > -rw-r--r--   1  root    root   0 Aug 12 17:30 cluster.vmsd
> >  /vmfs/volues/5208f99d-760cf4a2-000c29520788/cluster.vmdk #
>
> Okay, libvirt should not have created a directory named cluster.vmdk,
> I think there is a bug, but I can't test it at the moment because my
> ESX setup is unavailable due to a hardware failure.
>
> > Can anyone tell me how to start a node and why I am getting that error
> > message.
> > Below is my xml file.
> >
> > <domain type='vmware'>
> >   <name>testNode</name>
> >   <uuid>50115e16-9bdc-49d7-f171-53c4d7f91710</uuid>
> >   <memory>1048576</memory>
> >   <currentMemory>1048576</currentMemory>
> >   <vcpu>1</vcpu>
> >   <os>
> >     <type arch='x86_64'>hvm</type>
> >   </os>
> >   <devices>
> >     <disk type='file' device='disk'>
> >       <source file='[datastore1] cluster.vmdk'/>
>
> Here is a problem. Typically all files of a virtual machine should be
> located in a subdirectory named after the virtual machine. But you
> placed the VMDK file in the datastore root. I think this confused
> libvirt. This should read
>
> <source file='[datastore1] testNode/cluster.vmdk'/>
>
> instead. Also "virsh define" cannot create this VMDK file for you
if
> it doesn't exist already, because there is not enough information in
> the domain XML. You need to do this yourself using "virsh
vol-create".
>
> I suggest you delete the cluster.vmdk directory from the datastore and
> start over with the domain XML modified as suggested, but create
> "[datastore1] testNode/cluster.vmdk" before defining the virtual
> machine.
>
> --
> Matthias Bolte
> http://photron.blogspot.com
>