[openstack-dev] [nova] The same SRIOV / NFV CI failures missed a regression, why?

Moshe Levi moshele at mellanox.com
Mon Mar 28 17:13:59 UTC 2016



> -----Original Message-----
> From: Matt Riedemann [mailto:mriedem at linux.vnet.ibm.com]
> Sent: Thursday, March 24, 2016 3:35 PM
> To: OpenStack Development Mailing List (not for usage questions) <openstack-
> dev at lists.openstack.org>
> Subject: [openstack-dev] [nova] The same SRIOV / NFV CI failures missed a
> regression, why?
> 
> We have another mitaka-rc-potential bug [1] due to a regression when
> detaching SR-IOV interfaces in the libvirt driver.
> 
> There were two NFV CIs that ran on the original change [2].
> 
> Both failed with the same devstack setup error [3][4].
> 
> So it sucks that we have a regression, it sucks that no one watched for those CI
> results before approving the change, and it really sucks in this case since it was
> specifically reported from mellanox for sriov which failed in [4]. But it happens.
> 
> What I'd like to know is, have the CI problems been fixed? There is a change up
> to fix the regression [5] and this time the Mellanox CI check is passing [6]. The
> Intel NFV CI hasn't reported, but with the mellanox one also testing the suspend
> scenario, it's probably good enough.
Patch-Set 6  of patch [2] passed in Mellanox CI  see http://144.76.193.39/ci-artifacts/262341/6/Nova-ML2-Sriov/testr_results.html.gz 
I am not sure why patch-set 7 failed. At first I thought it was because of that in  PS 6 we install oslo.utils==3.4.0 and in PS7 oslo.utils==3.5.0
but I could not find a difference that can be related to  this:
2016-02-16 05:05:42.164 7182 ERROR nova   File "/usr/local/lib/python2.7/dist-packages/oslo_utils/importutils.py", line 30, in import_class
2016-02-16 05:05:42.164 7182 ERROR nova     __import__(mod_str)
2016-02-16 05:05:42.164 7182 ERROR nova ValueError: Empty module name

Putting that a side,  the Mellanox CI in nova is currently running and passing so the SR-IOV for Ethernet is not broken.
The fix in [5] is for our SR-IOV InfiniBand solution. At the moment we only test it in neutron (SR-IOV InfiniBand solution) and 
the reason for that is that we don't have many physical server to run the CI  for nova and neutron. 

> 
> [1] https://bugs.launchpad.net/nova/+bug/1560860
> [2] https://review.openstack.org/#/c/262341/
> [3]
> http://intel-openstack-ci-logs.ovh/compute-
> ci/refs/changes/41/262341/7/compute-nfv-
> flavors/20160215_232057/screen/n-sch.log.gz
> [4]
> http://144.76.193.39/ci-artifacts/262341/7/Nova-ML2-Sriov/logs/n-sch.log.gz
> [5] https://review.openstack.org/#/c/296305/
> [6]
> http://144.76.193.39/ci-artifacts/296305/1/Nova-ML2-
> Sriov/testr_results.html.gz
> 
> --
> 
> Thanks,
> 
> Matt Riedemann
> 
> 
> ________________________________________________________________
> __________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



More information about the OpenStack-dev mailing list