[Libvirt-cim] Re: Test Run Summary (Nov 17 2009): KVM on Fedora release 12 (Constantine) with Pegasus

Deepti B Kalakeri deeptik at linux.vnet.ibm.com
Fri Nov 20 06:15:12 UTC 2009



Kaitlin Rupert wrote:
> Deepti B Kalakeri wrote:
>>
>>
>> Deepti B Kalakeri wrote:
>>> =================================================
>>> Test Run Summary (Nov 17 2009): KVM on Fedora release 12 
>>> (Constantine) with Pegasus
>>> =================================================
>>> Distro: Fedora release 12 (Constantine)
>>> Kernel: 2.6.31.5-127.fc12.x86_64
>>> libvirt: 0.7.1
>>> Hypervisor: QEMU 0.11.0
>>> CIMOM: Pegasus 2.9.0
>>> Libvirt-cim revision: 1009
>>> Libvirt-cim changeset: 33a6a50f64e0
>>> Cimtest revision: 794
>>> Cimtest changeset: c7561bf61126
>>> Total test execution: Unknown
>>> =================================================
>>> FAIL      : 4
>>> XFAIL     : 4
>>> SKIP      : 11
>>> PASS      : 156
>>> -----------------
>>> Total     : 175
>>> =================================================
>>> FAIL Test Summary:
>>> ElementAllocatedFromPool - 01_forward.py: FAIL
>>> ResourceAllocationFromPool - 01_forward.py: FAIL
>>> ResourceAllocationFromPool - 02_reverse.py: FAIL
>> There were two diskpool cimtest-diskpool and default on the machine 
>> accessing the same /var/lib/libvirt/images pool path.
>> Libvirt-CIM was returning association information containing default 
>> diskpool information than cimtest-diskpool information.
>> Hence the tests failed. When I removed the default pool on the 
>> machine and ran the test all of them passed.
>> I think libvirt should give an error when two pools are being created 
>> with the same path or should  Libvirt-CIM handle this?
>
> Were both pools active?  And they were both "dir" type pools, right?
>
> I could see libvirt allowing two pools to represent the same path if 
> only one of those pools is active at a given time. But if both are 
> active, it seems redundant and potentially confusing.
Yes exactly it becomes confusing for an end user if such a scenario 
occurs, where two pool having all the same parameters except for the 
names, except for the uuid's are allowed to create.
My strong feeling is that we need to handle this via an error message.


>
> I can bring this up on the libvirt list.

Heres the two dir type disk pools that were present on the system:

# virsh pool-list --all
Name                 State      Autostart
-----------------------------------------
cimtest-diskpool     active     no       
default              active     no       

[root at elm3b151 cimtest]# virsh pool-dumpxml cimtest-diskpool
<pool type='dir'>
  <name>cimtest-diskpool</name>
  <uuid>2f4c1b4d-9c60-193c-8988-3ff3695b3c73</uuid>
  <capacity>61662662656</capacity>
  <allocation>27965640704</allocation>
  <available>33697021952</available>
  <source>
  </source>
  <target>
    <path>/var/lib/libvirt/images</path>
    <permissions>
      <mode>0700</mode>
      <owner>0</owner>
      <group>0</group>
    </permissions>
  </target>
</pool>

# virsh pool-dumpxml default
<pool type='dir'>
  <name>default</name>
  <uuid>a4ccaa1a-7ed4-99ab-2ae6-e10d9e6acd0e</uuid>
  <capacity>61662662656</capacity>
  <allocation>27625598976</allocation>
  <available>34037063680</available>
  <source>
  </source>
  <target>
    <path>/var/lib/libvirt/images</path>
    <permissions>
      <mode>0700</mode>
      <owner>0</owner>
      <group>0</group>
    </permissions>
  </target>
</pool>


-- 
Thanks and Regards,
Deepti B. Kalakeri
IBM Linux Technology Center
deeptik at linux.vnet.ibm.com




More information about the Libvirt-cim mailing list