Re: [Libvirt-cim] cimtest results.
"Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> 10/27/2011 05:17 AM
Please respond to eblima@br.ibm.com
To
Sharad Mishra/Beaverton/IBM@IBMUS
cc
eblima@br.ibm.com, Chip Vincent/Raleigh/IBM@IBMUS, wayne <xiawenc@linux.vnet.ibm.com>, Gareth S Bestor/Poughkeepsie/IBM@IBMUS
Subject
Re: cimtest results.
On 10/26/2011 08:07 PM, snmishra@us.ibm.com wrote:
Hi,
I see bunch of tests failing after I get cimtest from upstream and apply the following patches -
https://www.redhat.com/archives/libvirt-cim/2011-October/msg00051.html https://www.redhat.com/archives/libvirt-cim/2011-October/msg00052.html https://www.redhat.com/archives/libvirt-cim/2011-October/msg00045.html
If I run cimtest from rev 879 (which is upstream minus latest 4
"Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> wrote on 10/27/2011 05:17:58 AM: patches)
and compare the results with the run using latest upstream plus above three patches I see following extra failures -
VirtualSystemManagementService - 08_modifyresource.py: FAIL VirtualSystemManagementService - 09_procrasd_persist.py: FAIL VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL VirtualSystemManagementService - 31_unset_netrasd.py: FAIL VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL KVMRedirectionSAP - 01_enum_KVMredSAP.py: FAIL
I understand that 31_* and 32_* are new tests and that is the reason I did not see them in my run with older rev. But we need to figure out why these tests are failing.
The buggy patch is 883 ([TEST] XenKvmLib: Add cdrom device description to domain) and it looks weird that you actually ran revision 879 as it should not include VSMS 31 and 32. Can you try again with revision 882?
When I wrote "older rev" above, I meant rev 879. I did not see 31_ and 32_ in rev 879 and that is understandable since those tests were added in later rev. I am trying to understand why above tests are failing with 883 + cimtest patches on list (the three I pointed out above)? -Sharad
Best regards, Etrunko
-- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
On 10/27/2011 12:54 PM, Sharad Mishra wrote:
On 10/26/2011 08:07 PM, snmishra@us.ibm.com wrote:
Hi,
I see bunch of tests failing after I get cimtest from upstream and apply the following patches -
https://www.redhat.com/archives/libvirt-cim/2011-October/msg00051.html https://www.redhat.com/archives/libvirt-cim/2011-October/msg00052.html https://www.redhat.com/archives/libvirt-cim/2011-October/msg00045.html
If I run cimtest from rev 879 (which is upstream minus latest 4 patches) and compare the results with the run using latest upstream plus above three patches I see following extra failures -
VirtualSystemManagementService - 08_modifyresource.py: FAIL VirtualSystemManagementService - 09_procrasd_persist.py: FAIL VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL VirtualSystemManagementService - 31_unset_netrasd.py: FAIL VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL KVMRedirectionSAP - 01_enum_KVMredSAP.py: FAIL
I understand that 31_* and 32_* are new tests and that is the reason I did not see them in my run with older rev. But we need to figure out why these tests are failing.
The buggy patch is 883 ([TEST] XenKvmLib: Add cdrom device description to domain) and it looks weird that you actually ran revision 879 as it should not include VSMS 31 and 32. Can you try again with revision 882?
When I wrote "older rev" above, I meant rev 879. I did not see 31_ and 32_ in rev 879 and that is understandable since those tests were added in later rev. I am trying to understand why above tests are failing with 883 + cimtest patches on list (the three I pointed out above)?
So you are saying that in rev 879 the other tests (08, 09 and 11) pass? And can you provide the log when they fail? Best regards, Eduardo. -- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
"Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> 10/27/2011 10:42 AM
Please respond to eblima@br.ibm.com
To
List for discussion and development of libvirt CIM <libvirt-cim@redhat.com>
cc
Sharad Mishra/Beaverton/IBM@IBMUS, eblima@br.ibm.com
Subject
Re: [Libvirt-cim] cimtest results.
On 10/27/2011 12:54 PM, Sharad Mishra wrote:
On 10/26/2011 08:07 PM, snmishra@us.ibm.com wrote:
Hi,
I see bunch of tests failing after I get cimtest from upstream and
apply
the following patches -
https://www.redhat.com/archives/libvirt-cim/2011-October/msg00051.html
https://www.redhat.com/archives/libvirt-cim/2011-October/msg00052.html
https://www.redhat.com/archives/libvirt-cim/2011-October/msg00045.html
If I run cimtest from rev 879 (which is upstream minus latest 4
Here are the short logs on failures. I would suggest running cimtest with latest upstream + patches on list and make sure that these patches are not breaking existing tests. In general, before posting any patch, complete cimtest should be run and it should be verified that new patch is not breaking any cimtest. ++++++++++++++++ VirtualSystemManagementService - 09_procrasd_persist.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - Unable to define procrasd_persist_dom InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - DefineSystem with (KiloBytes) units failed InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest VirtualSystemManagementService - 31_unset_netrasd.py: FAIL ERROR - Current 'virtio' and expected '' ResourceSubType differ VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL ERROR - New media '/dev/null' does not match expected '' ERROR - TypeError : 'NoneType' object is unsubscriptable Traceback (most recent call last): File "/home/cimtest/suites/libvirt-cim/lib/XenKvmLib/const.py", line 141, in do_try rc = f() File "32_modify_cdrom_media.py", line 223, in main inst = modify_media(cim, inst, media_path) File "32_modify_cdrom_media.py", line 114, in modify_media val = set_device_addr(inst, addr) File "32_modify_cdrom_media.py", line 97, in set_device_addr inst["InstanceID"], TypeError: 'NoneType' object is unsubscriptable ERROR - None VirtualSystemManagementService - 08_modifyresource.py: FAIL ERROR - (4, u'CIM_ERR_INVALID_PARAMETER: Missing InstanceID in RASD') ERROR - Error invoking ModifyRS: mod_disk_res ERROR - ModifyResourceSettings call failed InvokeMethod(ModifyResourceSettings): CIM_ERR_INVALID_PARAMETER: Missing InstanceID in RASD Sharad Mishra Open Virtualization Linux Technology Center IBM "Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> wrote on 10/27/2011 10:42:17 AM: patches)
and compare the results with the run using latest upstream plus above three patches I see following extra failures -
VirtualSystemManagementService - 08_modifyresource.py: FAIL VirtualSystemManagementService - 09_procrasd_persist.py: FAIL VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL VirtualSystemManagementService - 31_unset_netrasd.py: FAIL VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL KVMRedirectionSAP - 01_enum_KVMredSAP.py: FAIL
I understand that 31_* and 32_* are new tests and that is the reason I did not see them in my run with older rev. But we need to figure out why these tests are failing.
The buggy patch is 883 ([TEST] XenKvmLib: Add cdrom device description to domain) and it looks weird that you actually ran revision 879 as it should not include VSMS 31 and 32. Can you try again with revision 882?
When I wrote "older rev" above, I meant rev 879. I did not see 31_ and 32_ in rev 879 and that is understandable since those tests were added in later rev. I am trying to understand why above tests are failing with 883 + cimtest patches on list (the three I pointed out above)?
So you are saying that in rev 879 the other tests (08, 09 and 11) pass? And can you provide the log when they fail?
Best regards, Eduardo.
-- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
On 10/27/2011 04:25 PM, Sharad Mishra wrote:
Here are the short logs on failures.
I would suggest running cimtest with latest upstream + patches on list and make sure that these patches are not breaking existing tests. In general, before posting any patch, complete cimtest should be run and it should be verified that new patch is not breaking any cimtest.
Sorry, I will pay more attention on this from now on. I am using the latest HEAD plus patches I've sent to revert the changes on XenKvmLib/vxml.py. Those patches are in the branch for-review of my cimtest tree in gitorious. git://gitorious.org/~eblima/libvirt-cim/etrunko-cimtest.git
++++++++++++++++
VirtualSystemManagementService - 09_procrasd_persist.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - Unable to define procrasd_persist_dom InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest
In my case this one fails with a different message: ====09_procrasd_persist.py Log==== Limit is None, expected 512
VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - DefineSystem with (KiloBytes) units failed InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest
This one matches: ====11_define_memrasdunits.py Log==== Defining with Bytes = 2147483648 Verified 2097152 KB Defining with KiloBytes = 2097152 Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 DefineSystem with (KiloBytes) units failed This test worked fine until the patch that introduced the changes in XenKvmLib/vxml.py was integrated. What is really weird is that even reverting those changes it still fails. :/
VirtualSystemManagementService - 31_unset_netrasd.py: FAIL ERROR - Current 'virtio' and expected '' ResourceSubType differ
VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL ERROR - New media '/dev/null' does not match expected '' ERROR - TypeError : 'NoneType' object is unsubscriptable Traceback (most recent call last): File "/home/cimtest/suites/libvirt-cim/lib/XenKvmLib/const.py", line 141, in do_try rc = f() File "32_modify_cdrom_media.py", line 223, in main inst = modify_media(cim, inst, media_path) File "32_modify_cdrom_media.py", line 114, in modify_media val = set_device_addr(inst, addr) File "32_modify_cdrom_media.py", line 97, in set_device_addr inst["InstanceID"], TypeError: 'NoneType' object is unsubscriptable ERROR - None
31 and 32 expose a bug in pegasus version 2.10 and 2.11.0. Make sure you have either 2.9 or 2.11.1.
VirtualSystemManagementService - 08_modifyresource.py: FAIL ERROR - (4, u'CIM_ERR_INVALID_PARAMETER: Missing InstanceID in RASD') ERROR - Error invoking ModifyRS: mod_disk_res ERROR - ModifyResourceSettings call failed InvokeMethod(ModifyResourceSettings): CIM_ERR_INVALID_PARAMETER: Missing InstanceID in RASD
This one works for me.... :/ Would be nice to have results from the others as well. I am investigating and will post a fix asap. Sorry for -- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
No problem Eduardo. In my eagerness to close a defect I have posted patches without first running all cimtests. So, this was a reminder to me too. Regards, Sharad Mishra Open Virtualization Linux Technology Center IBM "Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> wrote on 10/27/2011 01:16:38 PM:
"Eduardo Lima (Etrunko)" <eblima@linux.vnet.ibm.com> 10/27/2011 01:16 PM
Please respond to eblima@br.ibm.com
To
Sharad Mishra/Beaverton/IBM@IBMUS
cc
eblima@br.ibm.com, List for discussion and development of libvirt CIM <libvirt-cim@redhat.com>
Subject
Re: [Libvirt-cim] cimtest results.
On 10/27/2011 04:25 PM, Sharad Mishra wrote:
Here are the short logs on failures.
I would suggest running cimtest with latest upstream + patches on list and make sure that these patches are not breaking existing tests. In general, before posting any patch, complete cimtest should be run and it should be verified that new patch is not breaking any cimtest.
Sorry, I will pay more attention on this from now on. I am using the latest HEAD plus patches I've sent to revert the changes on XenKvmLib/vxml.py. Those patches are in the branch for-review of my cimtest tree in gitorious.
git://gitorious.org/~eblima/libvirt-cim/etrunko-cimtest.git
++++++++++++++++
VirtualSystemManagementService - 09_procrasd_persist.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - Unable to define procrasd_persist_dom InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest
In my case this one fails with a different message:
====09_procrasd_persist.py Log==== Limit is None, expected 512
VirtualSystemManagementService - 11_define_memrasdunits.py: FAIL ERROR - Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 ERROR - DefineSystem with (KiloBytes) units failed InvokeMethod(DefineSystem): CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest
This one matches:
====11_define_memrasdunits.py Log==== Defining with Bytes = 2147483648 Verified 2097152 KB Defining with KiloBytes = 2097152 Got CIM error CIM_ERR_FAILED: ResourceSettings Error: VirtualDevice property must be unique for each DiskResourceAllocationSettingData in a single guest with return code 1 DefineSystem with (KiloBytes) units failed
This test worked fine until the patch that introduced the changes in XenKvmLib/vxml.py was integrated. What is really weird is that even reverting those changes it still fails. :/
VirtualSystemManagementService - 31_unset_netrasd.py: FAIL ERROR - Current 'virtio' and expected '' ResourceSubType differ
VirtualSystemManagementService - 32_modify_cdrom_media.py: FAIL ERROR - New media '/dev/null' does not match expected '' ERROR - TypeError : 'NoneType' object is unsubscriptable Traceback (most recent call last): File "/home/cimtest/suites/libvirt-cim/lib/XenKvmLib/const.py", line 141, in do_try rc = f() File "32_modify_cdrom_media.py", line 223, in main inst = modify_media(cim, inst, media_path) File "32_modify_cdrom_media.py", line 114, in modify_media val = set_device_addr(inst, addr) File "32_modify_cdrom_media.py", line 97, in set_device_addr inst["InstanceID"], TypeError: 'NoneType' object is unsubscriptable ERROR - None
31 and 32 expose a bug in pegasus version 2.10 and 2.11.0. Make sure you have either 2.9 or 2.11.1.
VirtualSystemManagementService - 08_modifyresource.py: FAIL ERROR - (4, u'CIM_ERR_INVALID_PARAMETER: Missing InstanceID in RASD') ERROR - Error invoking ModifyRS: mod_disk_res ERROR - ModifyResourceSettings call failed InvokeMethod(ModifyResourceSettings): CIM_ERR_INVALID_PARAMETER:
Missing
InstanceID in RASD
This one works for me.... :/ Would be nice to have results from the others as well.
I am investigating and will post a fix asap. Sorry for
-- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
On 10/27/2011 06:29 PM, Sharad Mishra wrote:
No problem Eduardo. In my eagerness to close a defect I have posted patches without first running all cimtests. So, this was a reminder to me too.
I have just submitted a patch to fix 09 and 11. The patch is also in the for-review branch in my gitorious tree. Please let me know if there are any other issues left. Best Regards,
git://gitorious.org/~eblima/libvirt-cim/etrunko-cimtest.git
-- Eduardo de Barros Lima Software Engineer, Open Virtualization Linux Technology Center - IBM/Brazil eblima@br.ibm.com
participants (2)
-
Eduardo Lima (Etrunko) -
Sharad Mishra