Connections default to 'public' interface when attempting to run
openstack commands. This breaks setups where the network hosts are
separate from the controllers.
This change adds an ``openstack_interface`` variable to select the
correct endpoint to use, which defaults to ``internal``.
Co-Authored-By: Michael Senizaiz <michael@r-hpc.com>
Change-Id: Ifa766d2cc3ed7077f03c571398072ad5117701c6
Story: 2006814
Task: 37374
Hopefully this provides access to relevant information without breaking
the flow.
Change-Id: I40b71cdf782dda6971dafb5894670c30e446ccb0
Story: 2004337
Task: 37387
Diskimage builder supports specifying a list of packages to install via
-p. We currently support this for the root image, but not IPA images.
This change adds a new configuration option, 'ipa_build_dib_packages',
that should be a list of additional packages to install in locally built
IPA images. This affects the following commands:
kayobe seed deployment image build
kayobe overcloud deployment image build
This depends on
https://github.com/stackhpc/ansible-role-os-images/pull/25.
Also adds an example to the documentation for the equivalent option for
root image.
Change-Id: I04191d9541894b6a264e966c9ecb1056e0edade4
Story: 2006855
Task: 37446
The kolla-ansible CLI now requires the kolla_ansible python module to be
importable. We need to activate the virtualenv to make this work.
Change-Id: I43d4761409df3eea146dc7ff16b4c8d80855a91b
We no longer need to cd to the kayobe repo path, since all tested
versions of kayobe support pip installation.
Change-Id: Id7b574cbc6bb5195eee16cfc8443075712ce3254
Ideally we will fix these issues, but documenting them is a good first
step.
Change-Id: I9a840005d62b974f0681ac39e1ff02ae3ff0120e
Story: 2004967
Task: 29410
Also reorder upgrade documentation to make it clear that configuration
migration is something that should be done as a preparatory task.
Change-Id: I97ada3c6fd94a9ab7b5add97e90717f2a56bc8f3
Story: 2006677
Task: 36953
This variable is mentioned in "Host Configuration" but was missing from
the Docker extra-vars file.
Change-Id: I22cf878c99824372adc17582f9bbba2db98b3eb0
If the kayobe overcloud provision command is executed with overcloud
ironic nodes without names, cloud-init will fail with the following
Traceback when the deployed image boots:
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/cloudinit/sources/__init__.py", line 447, in find_source
if s.get_data():
File "/usr/lib/python2.7/site-packages/cloudinit/sources/__init__.py", line 132, in get_data
self._get_standardized_metadata())
File "/usr/lib/python2.7/site-packages/cloudinit/sources/__init__.py", line 110, in _get_standardized_metadata
'local-hostname': self.get_hostname(),
File "/usr/lib/python2.7/site-packages/cloudinit/sources/__init__.py", line 317, in get_hostname
if util.is_ipv4(lhost):
File "/usr/lib/python2.7/site-packages/cloudinit/util.py", line 544, in is_ipv4
toks = instr.split('.')
This is a bug [1] in cloud-init. The symptom is that hosts provision
successfully, but are not accessible via SSH. A solution is to ensure
that all nodes are named.
Until we have a release of CentOS with the cloud-init fix included, we
should at least mention this limitation in our documentation.
[1] https://bugs.launchpad.net/cloud-init/+bug/1852100
Change-Id: If54f56fb9f0b626d9fae80db622b0feeae5464b9
Story: 2006832
Task: 37405
Ironic now provides DIB-built CentOS images at
https://tarballs.openstack.org/ironic-python-agent/dib/files. However,
if Bifrost is configured to use these, then the 'kayobe seed service
deploy' command fails because the checksum files are not available.
This change fixes the issue by allowing
'kolla_bifrost_ipa_kernel_checksum_url' and
'kolla_bifrost_ipa_ramdisk_checksum_url' to be overridden to empty
strings, indicating that there is no checksum.
Very similar to the issue fixed in https://review.opendev.org/688359 for
the overcloud.
Change-Id: If73cebb95c7d0596dc8087112fc19398a113f0cc
Story: 2006715
Task: 37407
1. Sphinx 1.8 introduced [1] the '--keep-going' argument which, as its name
suggests, keeps the build running when it encounters non-fatal errors.
This is exceptionally useful in avoiding a continuous edit-build loop
when undertaking large doc reworks where multiple errors may be
introduced.
2. Cleanup the folders before run commands
[1] https://github.com/sphinx-doc/sphinx/commit/e3483e9b045
Change-Id: Ib3133820a3d3ca3027ea2f9bc9b6907929cff8ed
While not proper documentation, this should help people follow the ZTP
model.
Change-Id: If84a472826110bce151109ce80e70e1cef2a8177
Story: 2006640
Task: 37389
If you run 'kayobe overcloud provision' for the first time on a system,
and do not include the seed in the limit (e.g. --limit overcloud), then
the following error is seen:
Destination directory /etc/kolla/bifrost/inventory/host_vars does
not exist
This is because the first play in ansible/kolla-bifrost-hostvars.yml
executes against the seed. This is not ideal, as the --limit argment is
useful with this command, as a way to provision specific nodes.
This change fixes the issue by moving the initialisation tasks to the
second play which is targeted at the overcloud group. The run_once
parameter ensures it happens once.
Change-Id: I7dff72a50f0592a3564c9d84d684db562c30626a
Story: 2006757
Task: 37247
Backport: Stein, Rocky, Queens
In https://review.opendev.org/#/c/671502/ kolla ansible virtualenv
creation was modified to only happen on hosts in the kolla-ansible
inventory. However, that change omitted the seed. It passed CI because
the seed and overcloud are both in the inventory.
This change adds the seed group to the inventory_hostname lookup.
Change-Id: Ib5fb726600debee5df1f7085ea37b583bfccefcf
Story: 2006247
Task: 37368
We don't run kolla-target-venv during `seed hypervisor host configure`.
This playbook is limited to seed:overcloud anyway.
Change-Id: Ia2c7c151aca75aadc1efc406383b3966d77d4b9e
