[Libvirt-cim] Re: Test Run Summary (Aug 27 2009): KVM on Fedora release 11 (Leonidas) with Pegasuslist

[Kaitlin Rupert]

>> =================================================
>> FAIL Test Summary:
>> ComputerSystemIndication - 01_created_indication.py: FAIL
>> ElementAllocatedFromPool - 01_forward.py: FAIL
> This is not a test case issue. The test failed because there were two 
> conflicting diskpool with the same target dir used while creating it.
> 

> The following query was failing to return cimtest-diskpool and instead 
> returned DiskPool/tmp because of which the test failed:
> 
> #wbemcli ain -ac KVM_ElementAllocatedFromPool  
> http://root:<passwd>@localhost:5988/root/virt:KVM_LogicalDisk.CreationClassName="KVM_LogicalDisk",DeviceID="hd_domain/hda",SystemCreationClassName="KVM_ComputerSystem",SystemName="hd_domain" 
> 
> <hostname>/root/virt:KVM_DiskPool.InstanceID="DiskPool/tmp"
> 
> 
> Once I deleted the tmp diskpool and reran the test it passed.
> #wbemcli ain -ac KVM_ElementAllocatedFromPool  
> http://root:<passwd>@localhost:5988/root/virt:KVM_LogicalDisk.CreationClassName="KVM_LogicalDisk",DeviceID="hd_domain/hda",SystemCreationClassName="KVM_ComputerSystem",SystemName="hd_domain" 
> 
> <hostname>/root/virt:KVM_DiskPool.InstanceID="DiskPool/cimtest-diskpool"
> 
> So, I think libvirt should have not allowed to create two pools which 
> used the same target path, or libvirt-cim will have to handle this 
> condition.

This is a really good point.  libvirt should check for this, but we 
could check as well.  I'll make note of it as something that needs to be 
fixed.

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