Add support for configuring apt's proxy setting on Ubuntu hosts.

Change-Id: Iea1daff70fca5cf49f4e7f44af71a900678bb5c9
Story: 2009035
Task: 42782

Change-Id: I4ed284173841def54953d832c3fcd288e1d2605a

Change-Id: I670593ba8cbcd7c523b9bd6c42b4d0ca29e856ea

Currently the logs get printed in Zuul job stdout and only a small portion
of them is redirected to a file - this patch changes that to send all output
to desired files.

Change-Id: I45e937ec7c102e89edee20679b8959bf7f6c7152

Arguments are passed through to kayobe-env in kayobe-config, which
allows to set the Kayobe environment.

Change-Id: I4c72e32e5379237340284a09874b0c500e41ad0f

This may be used to deploy different images on different hosts.

Change-Id: I941cc28a914f3a56f50abdde70f13d6616ff52e7
Story: 2002098
Task: 41694

Change-Id: I70d5f78871867f3a71150eacc3df1857949e6df9

Tox jobs are timing out in CI due to pip dependency resolution
backtracking taking too long. When run locally, pip fails to resolve
dependencies due to upper-constraints now allowing pluggy 1.0.0 [1],
while molecule still depends on pluggy >= 0.7.1, < 1.0. See similar
tripleo bug for more details [2].

Split molecule dependencies in a separate requirements file and remove
upper constraints from its tox environment for now. As a bonus, it
should speed up execution of the jobs due to installing fewer
dependencies.

Stop using upper-constraints when installing docker python library [3].
This part of the commit can be reverted when docker 5.0.2 is in upper
constraints.

Update Kolla inventory template following renaming of haproxy role to
loadbalancer [4]. With the existing template, haproxy/keepalived
containers are not created and service deployment fails while trying to
connect to MySQL using the VIP.

[1] https://review.opendev.org/c/openstack/requirements/+/806680
[2] https://bugs.launchpad.net/tripleo/+bug/1942508
[3] https://review.opendev.org/c/openstack/kayobe/+/807128
[4] https://review.opendev.org/c/openstack/kolla-ansible/+/770618



Change-Id: I4cad83c92c32d1db2a5d3c03a78a1c533a8b9633
Co-Authored-By: Mark Goddard <mark@stackhpc.com>

The CentOS Stream 8 section used the wrong heading style, which broke
the rest of the document, making the other headings the wrong sizes.

Change-Id: I8015c905074016c53c803ce40a299ad5c911ada6

The command `kayobe overcloud introspection data save` has been failing
due to lack of OpenStack credentials when running the bifrost dynamic
inventory. Add OS_CLOUD=bifrost.

Story: 2009129
Task: 43059
Change-Id: I4a15578135a48cf282672288f4db24ded65f3b82

By default, Ansible injects a variable for every fact, prefixed with
ansible_. This can result in a large number of variables for each host,
which at scale can incur a performance penalty. Ansible provides a
configuration option [0] that can be set to False to prevent this
injection of facts. In this case, facts should be referenced via
ansible_facts.<fact>.

This change updates all references to Ansible facts within Kayobe
from using individual fact variables to using the items in the
ansible_facts dictionary. This allows users to disable fact variable
injection in their Ansible configuration, which may provide some
performance improvement.

This change disables fact variable injection in the ansible
configuration used in CI, to catch any attempts to use the injected
variables.

[0] https://docs.ansible.com/ansible/latest/reference_appendices/config.html#inject-facts-as-vars

Story: 2007993
Task: 42464
Depends-On: https://review.opendev.org/c/openstack/kolla-ansible/+/791276

Change-Id: I14db53ed6e57d37bbd28dd5819e432e3fe6628b2

If no public API network is used, as in the case of typical all-in-one
environments, the checks in kolla-ansible.yml, task "Validate Kolla
Ansible API address configuration" may fail. This happens only when
using the new kolla_internal_vip_address variable, and not if the legacy
vip_address network attribute is used.

This change fixes the issue by only checking for valid external VIP
address/FQDN configuration when a separate public API network is in use.
If the external VIP/FQDN are not set, then Kayobe no longer defaults
their values to the internal VIP/FQDN, instead relying on the Kolla
Ansible defaults to do this. This may have consequences for anyone
relying on these variables in configuration.

Change-Id: I0b1c963006c044ade5c67c4046da5e2addc9a8c7
Story: 2009134
Task: 43070

Change-Id: I0c41d59123aba3351ac3b0a5a76cff517cca93e4

As per [1] mandatory configuration option is either baseurl, mirrorlist
or metalink.

EPEL repos use metalink and it's impossible to configure them via DNF
custom repos functionality right now.

[1]: https://docs.ansible.com/ansible/latest/collections/ansible/builtin/yum_repository_module.html#parameter-metalink

Change-Id: Ib267b648f011a7efa6440b6adf9ea8b9b5a0931f

Change-Id: I097b826cb477577fc9482bbc3b8875b71a505a99

Change-Id: Iecae3bf7d54ba0bfe8fbf68a3deae5f74f0b831b

Change-Id: I68332411e53c1d50ed56ddba3f46ee37db22f976

Change-Id: Ied01d2797420f844134dff8714e77a1ca6c1f7aa

Configuring veth pair MTU in the systemd-networkd .netdev files
does not set the correct MTU for the peer and there isn't
configuration option in the [Peer] section to do so [1].
This results in an MTU mismatch with peer having the default
(1500) value.

This change moves the MTU configuration from the .netdev file
to the [Link] section of the corresponding veth pair .network files.
Same approach is already recommended in [1] for the tun and
tap devices.

[1] https://www.freedesktop.org/software/systemd/man/systemd.netdev.html

Story: 2009072
Task: 42882
Change-Id: I082e1b807497d17ae40ddf86e1473fbfcdf625e2

Change-Id: I202e1d2ec54e8ebcddefab668f8e2486f1bb1aa6

The openstack-tox-docs job started failing with the following error:

    ! LaTeX Error: File `tgtermes.sty' not found.

This change adds the required font package to the bindep doc profile.

Change-Id: I9acf8fd03150b9cc03a011364f0c45929b9a9f87

Change-Id: I424c1e3e6c488913d316e6b792d429599ea4de7c

Documentation about DIB_DISABLE_KERNEL_CLEANUP was left in place after
the variable was removed in Ie25e88bb96f186399ae69837e93185127628ce86.

Backport: wallaby, victoria, ussuri

Change-Id: Iddb699aa98ddee8446d532ce4bd23ba9174fb38e

The iscsi deploy interface was removed from Ironic [1]. Switch to the
direct deploy interface instead, which is the default.

[1] https://review.opendev.org/c/openstack/ironic/+/789382

Depends-On: https://review.opendev.org/c/openstack/kolla-ansible/+/801267
Change-Id: Ia0dcd3d302e0f04adbda4f3abf75a0ca7dd20dee

The use case is that I want to set these credentials via extra vars
like this:

  kayobe overcloud container image build test -e kolla_docker_registry_username=$HARBOR_REGISTRY_USER -e kolla_docker_registry_password=$HARBOR_REGISTRY_PASSWORD

Supporting the empty string means that I don't need to use any
conditional logic.

Change-Id: Ie5c21597266c8e470994eba16f3c2ed2224d0dcb

Because regexes defined by Kayobe are not strict enough, Kolla can build
container images unrelated to the deployment. For example, the default
service configuration will build:

- centos-binary-barbican-keystone-listener (matches keystone)
- centos-binary-prometheus-memcached-exporter (matches memcached)

This commit makes regexes stricter to avoid unecessary builds. Note that
some images are still built unecessarily, such as nova-compute-ironic
when ironic is disabled or neutron-metadata-agent-ovn when ovn is
disabled.

Change-Id: I67f0e48ee03b150f6fe52b1a3e7e727b79207231

Migrating to Ironic hardware types was required in Rocky. This can now
be removed from upgrading instructions.

Change-Id: I0510db9fd1b8dd72540c83173267491b32728330