[Libvirt-cim] Re: XenFV on Pegasus Test Run Summary for Sep 04 2008

[Kaitlin Rupert]

>>
>>
>> The following are the test cases which are failing even when manually 
>> run and needs investigation:
>>
>> VirtualSystemManagementService - 06_addresource.py: FAIL
> This tc is failing ONLY on Xen/XenFV because of the following reasons:
> We are using an non-existing network name in the VS XML file, this works 
> for when defining the KVM guest, but for Xen/XenFV it fails.
>   Here is the debug messages:
>   XML file:
>       <domain type='xen'>
>       <uuid>30cde8f6-6581-4b1c-969c-04558ceff31c</uuid>
>       <name>rstest_domain</name>
> 
>    <on_poweroff>destroy</on_poweroff>
>    <on_crash>destroy</on_crash>
>    <os>
>      <type>linux</type>
>      <kernel>/tmp/default-xen-kernel</kernel>
>      <initrd>/tmp/default-xen-initrd</initrd>
>      <cmdline>TERM=xterm</cmdline>
>    </os>
>    <currentMemory>131072</currentMemory>
>    <memory>131072</memory>
>    <vcpu>1</vcpu>
>    <devices>
>    <interface type='bridge'>
>      <source bridge='testbridge'/>
>      <mac address='11:22:33:aa:bb:cc'/>
>      <script path='vif-bridge'/>
>    </interface>
>    <interface type='network'>
>      <mac address='99:aa:bb:cc:ee:ff'/>
>      <source network='cimtest-networkpool20'/>
>    </interface>
>    <disk type='file' device='disk'>
>      <source file='/tmp/default-xen-dimage'/>
>      <target dev='xvda'/>
>    </disk>
>    <disk type='file' device='disk'>
>      <source file='/tmp/default-xen-dimage.2ND'/>
>      <target dev='xvdb'/>     </disk>
>    </devices>
>    </domain>
> 
> 
>   _Virsh message:
> _    virsh define xen
>    libvir: QEMU error : no network with matching name
>    error: Failed to define domain from xen
>  
>   _Debug Messages:
> 
> _   misc_util.c(72): Connecting to libvirt with uri `xen'
>   libvir: QEMU error : no network with matching name
>   Virt_VirtualSystemManagementService.c(770): Failed to define domain 
> from XML
>   std_invokemethod.c(305): Method `AddResourceSettings' returned 1
> 
>   I believe the changes to this particular test case was done to avoid 
> the test case failing due to following error seen in the batch runs:
>   *"ERROR - Need to give different bridge name since it already exists   "
>  
>   *If yes, then I think we should revert back to the changes that was 
> prior to applying the patch *"[PATCH 4 of 6] [TEST] Make network related 
> changes to VSMS tests"*
>   Any other suggestions ?*

Yes, I'd agree with that. This test needs to create a network pool using 
the npool_name variable as the network name.  And make sure the pool is 
cleaned up and all that. That was my mistake - when I added the network 
pool creation to main.py, I forgot that this test needed an additional 
pool so it could add a different network device.

Thanks for catching this!

-- 
Kaitlin Rupert
IBM Linux Technology Center
kaitlin at linux.vnet.ibm.com