Related change: https://review.opendev.org/806562
Would need '--baremetal-deployment' to make overcloud
deploy provision the ControlPlane VIP.
Change-Id: Ife4b15a784c70fe031af98f3aa135737232d56a6
This review adds and extra-vars to
disable protected resourced types in
releases wallaby and after for
fs010, fs037, fs035 and fs020.
There is a need to disable the resource
registry override protection to merge:
https://review.opendev.org/798642
Closes-Bug: #1943579
Change-Id: I618c0a0d99c934fb65a6af30bc76096d52d4679d
featureset037 is
tripleo-ci-centos-8-scenario000-multinode-oooq-container-updates, as
well as the -(train|ussuri/victoria) variants. However ephemeral_heat
enablement is restricted to just the wallaby+ branches in the variable
definition.
This needs to be enabled explicitly for this job before we can proceed
with https://review.opendev.org/c/789930 to use ephemeral heat by
default.
Signed-off-by: James Slagle <jslagle@redhat.com>
Depends-On: I3d8a1282c388a8201ec1ce5c7d64796c0ce6026a
Change-Id: I81d5a8e86b1b2dd03d4ea29dee10b90efe460094
This patch removes:
* legacy tripleo-validations config boolean, since now the role
is in tripleo-validations repos and is enabling by another
configuration in the zuul config directly
* documentation, because its now outdated and a new one should
be moved under tripleo-validations repos
Change-Id: Ibc5e1aec08643d53760953ebb1c026e99689a2b6
This was added to reduce the CI wall time. As we're removing
overcloud plan, this won't work. We can skip the converge
step if required later if we want to reduce the time jobs
take.
Change-Id: Ifd0371274fcc42286217a5c47e3998aafa98ddcc
The variable run_tempest was used by validate-tempest. Since now we use
only os_tempest, this variable is no longer required, however, since
use_os_tempest is default to true, we must change it where the
run_tempest is false, like in the upgrade/update jobs, since we don't
run tempest on those.
Change-Id: Idb8166992976ce6303f95f72011aa65fe536931e
Ensure that the container_registry_file is referenced
correctly for upgrades involves releases stein
and before.
This is the third revision of patches to address
the related bug.
Change-Id: Ib036d9758f64ab78c6cf0f91c5fd788e0e167eda
Related-Bug: #1903581
This featureset runs with scenario000 who have only keystone installed,
and so, it's not possible to run tempest. Since we are moving
validate-tempest (which is already disabled) to os_tempest, we must
disable it here as well.
Change-Id: I223bcf16fd3184d33306fa84c09f55b0927f589c
Removes the redundant undercloud mapping from featuresets.
We've the default mapping in environments/undercloud.yaml.
Also this CI specific environment is redundant.
Change-Id: I72e41fb8cdebff12aab9704c2a9c4e368410c1ad
This option adds --debug to the cli command which generates an excessive
amount of logs. We shouldn't have this enabled by default.
Change-Id: I655aa888ca10cc622f80f3f267cdaddf844bdf81
From rocky onwards, the file used to identify the containers parameters
passed to be containers-prepare-parameter.yaml. This file had a different
format than in previous releases. At that point, a new variable was
introuduced in tripleo-upgrade to identify this containers parameter file:
uc_containers_prepare_file. To allow a transition from pre-rocky jobs to
post-rocky, the container_registry_file parameter was left to know the
name of the old containers prepare file. This file was removed from the
appended overcloud_deploy.sh files to avoid issues during the upgrade/
update prepare step [0].
Updates/Upgrades jobs were still using the container_registry_file as
the parameter used to set up the containers, pointing at the post-rocky
file containers-prepare-parameter.yaml. The code in [0] then removed it
from the appended -e environmetn files, running the overcloud update
prepare without the right containers-prepare-parameter.yaml.
[0] - https://review.opendev.org/#/c/717698
Change-Id: I6816db7382e7db6d0ae644c9487b2c5ee595a441
Closes-Bug: #1887176
fs037 and fs051, use config-download when running Queens jobs. The
config-download deployment uses tripleo-admin user for the ansible
execution, however we had hardcoded the zuul user (as Rocky, Stein
and master do use config-download by default). As a consequence, to
have the Queens jobs which use these featuresets working, we need
to specify the use of the overcloud-ssh-user to tripleo-admin.
Change-Id: I475c7c805fdcaa90b9b0183135d81f10faba3b8d
Closes-Bug: #1840763
Currently, we are using zuul user as overcloud_ssh_user during overcloud
deployment. While for upgrades, we are still passing tripleo-admin. This
is causing the upgrade prepare command to fail.
Change-Id: If4bd02dad8d69ba4c818ba8d800423bcc8ecfea4
Closes-Bug: #1813564
This normalizes the formatting of files from ci-scripts and config
folders in order to allow further hardening of linting.
Fixing linting was too big to be made in a single commit as it would
involve too many files to review and could merge conflicts with
existing changes. Thus doing it in few chunks would makes it possible.
Original full change is at https://review.openstack.org/#/c/627545/ and
will be the one merging the last.
Change-Id: Ifb215c0e2ea0ef7115897721f75ba8489bd59b97
The output from the containers image prepare
step is not the containers-default-parameters.yaml
file with the list of containers, but another file
named containers-prepare-parameters.yaml which
contains the ContainerImagePrepare parameter. This
parameter is used during deployment-steps to get
the list of containers.
Co-Authored-By: Quique Llorente <ellorent@redhat.com>
Closes-Bug: #1789881
Depends-On: If0c7786059f1d0d91d2279f45c67525908633378
Depends-On: I96109c2452d7568da4a9374661a8ba8b9ba71458
Change-Id: I9bd21b67a9bf9254458351e8ba784ed66f9949f4
Once [0] got merged, there is no chance that the
overcloud-upgrade role will run. Upgrade job have
all been moved to use the new tripleo-upgrade, so
we can dispense all overcloud-upgrade related
variables from the featuresets.
[0] - Iaf30244187d27909ab803175843056e0827d0b24
Change-Id: Ibe715cc4f604ba554db8dd4ab2d95391e4b1f311
fs020 is switched to containerized undercloud in [1],
it's failing since then. Switch to undercloud net config
to fix this. fs002 is fixed in [2].
Also change it in other containerized undercloud featuresets.
[1] https://review.openstack.org/#/c/572215/
[2] https://review.openstack.org/#/c/576987/
Change-Id: I992a49ef025065f4dc02c650105832e85d9eb8b8
Related-Bug: #1749707
We incorrectly dropped the config download items in
Ic5c5df350a7eb5291d45cf8e229af575fe9bed57 for featuresets that had it
enabled in queens. Since we won't be backporting the default changes in
tripleoclient back to queens we need to add a special case for these
featuresets in queens.
Change-Id: I9aa449a3198568c25d4dbc838669f493b0e3c103
Closes-Bug: #1771549
Now that config-download is the default across tripleoclient and
tripleo-heat-templates, we can update all the CI jobs to remove the
explicit --config-download and -e
environments/config-download-environment.yaml.
Note that --deployed-server didn't actually trigger any needed behaviors
with --config-download, so where it was used, it has now been removed as
well.
Change-Id: Ic5c5df350a7eb5291d45cf8e229af575fe9bed57
We discovered we need to put back stack updates into the converge
operations. This increases the run time of the upgrade procedure
considerably, and e.g. our voting CI job for updates might go over the
time limit. We'll need to start passing --update-plan-only to the
converge operation to keep CI on previous run times.
Change-Id: Idc5dbebc116aba2420d275aab3981186511f298e
Related-Bug: #1768586
As part of the upgrades/updates playbook refactor
it has been removed the existence of a 'legacy'
and 'new' upgrades playbook, which was being
controlled by the 'use_tripleo_upgrade' variable
Iaf30244187d27909ab803175843056e0827d0b24.
Also, it's been moved the parameter container_registry_file
from the overcloud update's playbook into its
corresponding fs.
Related-Bug: #1749740
Depends-On: Iaf30244187d27909ab803175843056e0827d0b24
Depends-On: Iec42b574d697f3acfd668636e3dd11217a3ea25e
Change-Id: I27f13b164bb78619f8830b875ec9f04445ce1e26
* https://review.openstack.org/551441 introduced test_black_regex
in order to skip test directly from featureset files. In order
to avoid confusion between test_regex and test_black_regex, we
are renaming test_regex to test_white_regex.
* It will fix all the featuresets.
Depends-On: https://review.openstack.org/560006
Change-Id: If66225c2e2f8c571cc85aebacee01431e4f64dcc
First we switched the user to tripleo-admin [1] which made CI on
updates job green, but broke Nova-based environments. We're now
switching the default back to heat-admin [2] which will break updates
job again, so we add support to tripleo-upgrade [3] to customize the
user name and set it here to the value for CI (deployed server
environments).
Fear not, when we switch to --config-download as our only supported
deployment method, tripleo-admin should become unconditionally present
in all environments, and we'll be able to converge to a single method
of logging into the overcloud nodes.
[1] I1ec66886d6d61ccf528e5e27872cf68c7a603727
[2] I0df57002b2305c1e2504c9f7a7d0c326d8ffcaf7
[3] Ib486a812aa9fd96bc753eda122cd05682d97b209
Co-Authored-By: Jose Luis Franco <jfrancoa@redhat.com>
Change-Id: I063d438fbeeded28134658aaefc6c8b7119aee43
Related-Bug: #1759845
Depends-On: I0df57002b2305c1e2504c9f7a7d0c326d8ffcaf7
In order to run some of the needed
tripleo-upgrade tasks, we need to set
tripleo_ci variable, as the role is being
used for downstream as well as upstream
testing and this is the flag to run specific
TripleO CI tasks.
Change-Id: I15530dce10d3e2bd08f69e6d71ca0ba95a27b30c