In test_create_server_from_volume_snapshot tearDown would trigger
bootable volume and snapshot deletion in the correct order but not
wait for the bootable volume to be deleted, causing snapshot
deletion to fail.
This change makes the scenario test explicitly delete the bootable
volume to avoid the race condition in tearDown.
Change-Id: I3fe2ce349ce629bfc96f6ddfb45b22687a1be137
Co-Authored-By: Alan Bishop <abishop@redhat.com>
Closes-Bug: 1863750
devstack-plugin-ceph-tempest is py2 job and not
supposed to run on master gate. Moving
devstack-plugin-ceph-tempest-py3 from experimental pipeline
to check.
Change-Id: I8a15d919f9902315c4a8d27d73dabbcc5be8d8de
If live migration tests are ran on environment with multiple AZ and
destination host for live migration is chosen from different AZ then
nova scheduler would stop such migration and live migration tests
would fail so in the case of CONF.compute.compute_volume_common_az
being set in config file, choose destination host from the same
availability zone (the same happens with cinder AZ).
Change-Id: I6bfad96ae3aa6cd8efd7ae2a0ce01c53140f617a
'test_delete_saving_image' fails because image
snapshot is produced quicker than tempest being able
to catch the 'SAVING' transition.
Hence, the testcase needs following refinement:-
1. testcase should catch the state of image in 'SAVING'
state only and then delete the image.
By this case, we can get the idea that image is
deleted in 'SAVING' state only and the testcase will
achieve its motto.
2. If the image trasits to a state other than
'SAVING', testcase can just delete the image.
Related-Bug: #1713163
Signed-off by: Soniya Vyas<svyas@redhat.com>
Change-Id: I50606c67e9f626935878aaa61d0adbc236272054
I saw in Neutron queue that default 2h timeout for this job wasn't
enough sometimes. So it's better to keep this job running a bit longer
sometimes rather than recheck everything.
Change-Id: I7c256f1ca5790cd13411d3af78ec97efcfd28b0a
spice_console and rdp_console are no longer used after
9f5adf8053, this is to deprecate
them for removal.
Change-Id: I8a882878493e3fe39bbfe0628991ee551c15d1b0
Create_test_server in test_resize_volume_backed_server_confirm didn't
use validation_resources nor validatable=True, which resulted in an
SSHTimeout error when using fixed IP for connect_method.
When using connect_method=floating, the validation check refers
to the server created in setUp instead of this one, which prevented
this bug from being caught.
Change-Id: I778236f456ef91e78a592f99e552cb3088b1e5db
Co-authored-by: zhufl <zhu.fanglei@zte.com.cn>
Closes-Bug: #1751418
This change replaces the attach_volume cleanup call to
wait_for_volume_resource_status with wait_for_volume_attachment_remove
for multiattach volumes. This newer waiter method being able to handle
cases where multiattached volumes remain attached and thus marked as
'in-use' after being detached from another instance.
Closes-Bug: #1858841
Change-Id: I4e9d253cad1b797940c2c1f922b66602ba592416
Avoid creating a SSH client with a proxy client having the same
connection parameters. This will prevent from creating a SSH tunnel
within the same client.
This patch also provides aditional information in the exception
message, in case of SSH connection error, about the proxy client
configured in the SSH client.
Change-Id: I4fcffb7e26b0adc0e39dc13085f93db6495c3bd6
Related-Bug: #1860586
Allowing BaseV2ComputeTest based test classes to specify
volume|placement_min_microversion and volume|placement_max_microversion.
Change-Id: If69f0456dc40b923b550bc648caaf0cb1a16e086
Stable branches till stable/rocky is using python
version <py3.6. Tempest test those branch in venv
but Tempest tox use the master upper-constraint[1]
which block installation due to dependencies
require >=py3.6. For exmaple, oslo.concurrency 4.0.0
is not compatible for <py3.6.
As we pin Tempest for EM stable brach, we should be
able to use stable constraint for Tempest installation
as well as while running during run-tempest playbook.
tox.ini is hard coded to use master constraint[1] which force
run-tempest to recreate the tox env and use the master constraint.
Devstack can set stable u-c to use via env var but this is
not enough as Tempest role run-tempest recreate the tox
env with master u-c.
- https://review.opendev.org/#/c/705089/4/lib/tempest
Currently we are fixing run-tempest to use stable constraint
till stable/rocky which use distro having <py3.6.
Ideally, we can change compatible branch constraint in tox.ini
while doing release tag for Tempest. But that can be done
for future tags only.
[1] bc9fe8eca8/tox.ini (L14)
Change-Id: I4268049ff91e5527e4b95fd8a70fa53ba27d56ed
test_tagged_attachment was failing when Tempest was set to connect
via fixed IP. The test was trying to attach an interface to an
instance that was still in the building state. Adding a 'wait_until'
to the method creating said instance fixes the issue.
Change-Id: Ia1252a22c54e3a84a8e5e779ed6ea5ee593760da
Closes-Bug: #1860680
If3ff71dbe636ef98072b15f6599e6e641c222f30 had switched this to python3.6
but this breaks on distros where 3.6 is no longer provided by default,
such as Fedora.
Closes-Bug: #1861308
Change-Id: I0e0b9e88eea2499cb1b9727a7787c050c869c819
Tempest, its plugins and their dependencies in ussuri
cycle are python-requires>=py3.6 which makes distro not
having python 3.6 to do hack to install py3.6 etc.
We cannot do >py3.6 in basepython so one way is to drop the
py3.5 support and basepython to py3 so that it can choose
the available py version in env.
It time to drop the py3.5 from Tempest. Last supported version of
Tempest for py3.5 is 23.0.0.
oslo drop py2 so we need to remove the py2 unit test job which
is failing and blocking gate.
There is seoarate work to drop the py2 things from Tempest in
compplete set.
Change-Id: I872dbb57df46d1193c0d409bb00d56f372d295fe
Related-Bug: 1861308
As per tempest design, if tempest conf is not having any region for
a service then identity region should be used. In case of placement
the default value is "RegionOne" which is considered as placement
region if region missing in tempest conf. In order to have identity
region to be used as default we need to change to empty string for
placement service. Empty string can be seen being used in other
services like volume, image etc.
Please check the below link for the logic of default region
https://github.com/openstack/tempest/blob/master/tempest/config.py#L1435
Change-Id: I2a1045cd0f9eb830493b2b4287ccee46223f3c59
The response data of test_routers_client are edited in order to be
consistent with the Reference API document [1]
[1] https://docs.openstack.org/api-ref/network/v2/index.html
Closes-Bug: 1838005
Change-Id: I9f5e57a3114ac047f87762def08b165bf377b600
This option duplicates the more generic
[compute-feature-enabled]/block_migration_for_live_migration now that
MIN_LIBVIRT_VERSION is >= 1.2.17 on all branches from stable/rocky. We
can deprecate this now ahead of removal once stable/pike and
stable/queens are no longer supported.
Change-Id: I8d8ae5ebf8e1d383f184527815f0a0c0ba9a343d
This change adds a specific waiter for tracking the removal of
attachments from a volume. This should allow callers to track the
detaching of multiattached volumes from instances while the volume
remains in an `in-use` state.
Related-bug: #1858841
Change-Id: I85917f683d2bc81fdf3f6d3351bf56fdb649740e
Use of os-tenant-name or OS_TENANT_NAME is deprecated.
Instead of it, os-project-name is expected to be used.
This patch depends on below mentioned patch, since
devstack still uses the deprecated argument.
Depends-on: I624e1dc57a3d3533322fb298c01f70241d0400ed
Signed-off by: Soniya Vyas<svyas@redhat.com>
Change-Id: I3b926f8abb6fce3d2eab93310ce3959c68aa9a6d
The review drops usage of skip_path() filter in the related tests
and uses raw_request() instead.
Normally a swift url is organised as host:port/info and
host:port/v1/AUTH_<tenant-id>, see
https://docs.openstack.org/api-ref/object-store/
But RadosGW API is organised as host:port/swift/info and
host:port/swift/v1/AUTH_<tenant-id>, see
https://docs.ceph.com/docs/master/radosgw/config-ref/#swift-settings
Close-bug: 1799981
Change-Id: I6a932639a05defe0f04c600afcc35a19662937af