Hi all, in preparation for the next stable/grizzly release (target release date is Aug 8, freeze week before that) we should check that periodic grizzly jobs are in good shape: * Horizon Recent horizon failures due to django-openstack-auth/keytoneclient updates[1] should be fixed when https://review.openstack.org/35309 is merged (the same fix is needed for stable/folsom) * Glance https://jenkins.openstack.org/job/periodic-glance-python27-stable-grizzly/se... to be failing randomly on integration testing and from talking to eglynn, fixing that is not going to be backportable, so we're going to have to live with that Also, stable/folsom which is in security-fixes-only mode, should be kept in shape: * Horizon - see above * Nova py26 https://jenkins.openstack.org/job/periodic-nova-python26-stable-folsom/shoul... be good now that eventlet released 0.13.0 with the fix for RHEL's python[2] (other branches carried a patch[3] which could be removed now) * Tempest https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-stable-folsom... https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-cinder-stable... seem to be failing on git_clone neutronclient, https://github.com/openstack-dev/devstack/commit/d68d6d1ee67e48b9a34e8209ab3... not enough on stable/folsom? Could infra folks have a look? Cheers, Alan [1] more info in horizon meeting log http://eavesdrop.openstack.org/meetings/horizon/2013/horizon.2013-07-09-22.0... [2] https://bitbucket.org/eventlet/eventlet/issue/89/add-a-timeout-argument-to-s... [3] https://github.com/openstack/nova/blob/stable/grizzly/contrib/redhat-eventle...
Hi Alan, "Alan Pevec" <apevec@gmail.com> wrote:
in preparation for the next stable/grizzly release (target release date is Aug 8, freeze week before that) we should check that periodic grizzly jobs are in good shape:
* Horizon Recent horizon failures due to django-openstack-auth/keytoneclient updates[1] should be fixed when https://review.openstack.org/35309 is merged (the same fix is needed for stable/folsom)
The fix has been merged in Grizzly and is now backported to Folsom at https://review.openstack.org/#/c/36664/ , although it's hitting the Tempest issues brought up below. Regards, Julie
[SNIP]
* Tempest
https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-stable-folsom...
https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-cinder-stable... seem to be failing on git_clone neutronclient, https://github.com/openstack-dev/devstack/commit/d68d6d1ee67e48b9a34e8209ab3... not enough on stable/folsom? Could infra folks have a look?
Cheers, Alan
[1] more info in horizon meeting log http://eavesdrop.openstack.org/meetings/horizon/2013/horizon.2013-07-09-22.0...
2013/7/11 Julie Pichon <jpichon@redhat.com>:
The fix has been merged in Grizzly and is now backported to Folsom at https://review.openstack.org/#/c/36664/ , although it's hitting the Tempest issues brought up below.
This review has been merged and folsom jobs are all clear now, thanks Julie! I'm looking forward to receive no bitrot emails tomorrow :) Cheers, Alan
2013/8/1 Alan Pevec <apevec@gmail.com>:
2013/7/11 Julie Pichon <jpichon@redhat.com>:
The fix has been merged in Grizzly and is now backported to Folsom at https://review.openstack.org/#/c/36664/ , although it's hitting the Tempest issues brought up below.
This review has been merged and folsom jobs are all clear now, thanks Julie! I'm looking forward to receive no bitrot emails tomorrow :)
I spoke too soon, today both horizon folsom and grizzly failed, comparing pip freeze output between https://jenkins.openstack.org/job/periodic-horizon-python27-stable-grizzly/1... and https://jenkins.openstack.org/job/periodic-horizon-python27-stable-grizzly/1... -django-openstack-auth==1.0.11 +django-openstack-auth==1.1.1 Julie, I need to ask for your help again, thanks in advance :) Cheers, Alan
"Alan Pevec" <apevec@gmail.com> wrote:
2013/8/1 Alan Pevec <apevec@gmail.com>:
2013/7/11 Julie Pichon <jpichon@redhat.com>:
The fix has been merged in Grizzly and is now backported to Folsom at https://review.openstack.org/#/c/36664/ , although it's hitting the Tempest issues brought up below.
This review has been merged and folsom jobs are all clear now, thanks Julie! I'm looking forward to receive no bitrot emails tomorrow :)
I spoke too soon, today both horizon folsom and grizzly failed, comparing pip freeze output between https://jenkins.openstack.org/job/periodic-horizon-python27-stable-grizzly/1... and https://jenkins.openstack.org/job/periodic-horizon-python27-stable-grizzly/1...
-django-openstack-auth==1.0.11 +django-openstack-auth==1.1.1
Julie, I need to ask for your help again, thanks in advance :)
Hi Alan, My bad, the initial requirement change should have been made more restrictive in the backports. Grizzly: https://review.openstack.org/#/c/39876/ Folsom: https://review.openstack.org/#/c/39879/ Cheers, Julie
Cheers, Alan
On 2013-07-11 14:37:29 +0200 (+0200), Alan Pevec wrote: [...]
* Tempest https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-stable-folsom... https://jenkins.openstack.org/job/periodic-tempest-devstack-vm-cinder-stable... seem to be failing on git_clone neutronclient, https://github.com/openstack-dev/devstack/commit/d68d6d1ee67e48b9a34e8209ab3... was not enough on stable/folsom? Could infra folks have a look?
Digging in a little, it looks like this is failing because openstack-infra/devstack-gate:devstack-vm-update-image.py is locally caching /opt/stack/new/python-neutronclient and then the stable/folsom branch of openstack-dev/devstack:stack.sh is looking for a copy at /opt/stack/new/python-quantumclient instead (since it refuses to download it when running in offline mode). I can see a few possible fixes to this, but I'll take a stab at just addressing it directly in stack.sh where it's being hard-coded and seeing if that works as expected. -- Jeremy Stanley
On 2013-07-11 16:35:28 +0000 (+0000), Jeremy Stanley wrote: [...]
I'll take a stab at just addressing it directly in stack.sh where it's being hard-coded and seeing if that works as expected.
So... while it seems to solve the issue at hand, I think it merely exposed new bitrot: https://bugs.launchpad.net/bugs/1200336 -- Jeremy Stanley
participants (3)
-
Alan Pevec
-
Jeremy Stanley
-
Julie Pichon