* Updates API version to 1.85 to permit an ``unhold`` verb
* Adds the ``deploy hold`` and ``clean hold`` provision states
to the internal state machine.
* Adds on documentation on steps to help provide greater clarity
to Ironic's users on how to utilize steps. It should be noted
this documentation also includes the power state reserved step
names from the DPU functionality patch.
* Fixes the state machine diagram. Changes type to PNG as SVG
rendering is broken due to python libraries utilized for SVG
generation which do not work on more recent Python versions.
Disable irmc virtual media test_prepare_instance_with_secure_boot
Explicitly close out test connection
When creating a test database, we should follow the same
pattern we know to be good, where an orphaned handler is not
left in memory to close out a connection.
Change connection style in dbTestBase to mirror how we do
database connections so they close out when we are done
Change migrations timeout to be >60 seoncds
In local testing, I found the migrations tended to take an
average of 70 seconds. Granted, my test machine is old, and slow
but the performance is very similar to a busy cloud provider.
As such, increase the timeout to a larger value so we can enable
the double migration test again.
Also use BASE_TEST_TIMEOUT as time limit for unit tests, failing
hard if that's passed.
Co-Authored-By: Julia Kreger <email@example.com>
All database migration testing in opestack is done through
an opportunistic worker model, where if the database is available
and correctly configured for testing, i.e. openstack-citest user
and access appropriately granted, then the tests will create and
However, this has been problematic with mysql as of recent, as we
have seen a long standing migration issue boil to the surface often
As a result, we're isolating that test down to it's own job so we
can limit the blast damage. This also helps us isolate is it all
of the tests, or is it just soley isolated down to the mysql test
run class, which is an additional data point.
By default, we continue to run Postgres migration tests in the
main jobs, as they haven't been impacted by this issue.
The default test runner behavior is to distribute various classes
across many runners, which is fine. But when you have lots of setup
to ensure an environment is correct for tests in a class, you
generally want to execute that together instead of separately.
As such, set the --parallel-class execution option for stestr.
Some tests tool and IDE may create a .local directory inside the repo
with virtualenvs for dependencies, other tools may create . directories
or files for temporary reports.
While they can be removed in a second time, or configured differently,
it's advisable to just exclude all file starting with . from the
flake8 tests to avoid confusion and possible unexpected errors.
- Remove skipsdist that it was never supported and causes breakage
when used with usedevelop.
- add script to allowlist for pep8 test
- disable setuptools autodiscovery
- Increase base VM memory according to new requirements for CS9
* Changed common exception imports from SQLAlchemy for ORM
query types which are now originated from the main exception
* Changed base join option usage to use objects instead of labels,
and defaulted all multi-row result sets to return data using
"selectinload" as opposed to operating with a join query to
avoid need to de-duplicate all result sets.
* Changed DeployTemplates to utilize objects instead of
field names for queries, and updated the associated join ORM
model's relationship record between DeployTemplate and
* Changed Ports, Chassis, Conductor, Volume Target/Connector
queries to lean towards use of select/update/delete queries as
opposed to ORM queries. Most of these changes revolved around
references of field names as opposed to model objects.
* This change also labels a few lines as "noqa", which is a result
of the style check rules getting triggered on statements as
needed for SQLAlchemy.
One of the major changes in SQLAlchemy 2.0 is the removal
of autocommit support. It turns out Ironic was using this quite
aggressively without even really being aware of it.
* Moved the declaritive_base to ORM, as noted in the SQLAlchemy 2.0
* Console testing caused us to become aware of issues around locking
where session synchronization, when autocommit was enabled, was
defaulted to False. The result of this is that you could have two
sessions have different results, which could results on different
threads, and where one could still attempt to lock based upon prior
information. Inherently, while this basically worked, it was
also sort of broken behavior. This resulted in locking being
rewritten to use the style mandated in SQLAlchemy 2.0 migration
documentation. This ultimately is due to locking, which is *heavily*
relied upon in Ironic, and in unit testing with sqlite, there are
no transactions, which means we can get some data inconsistency
in unit testing as well if we're reliant upon the database to
precisely and exactly return what we committed.
* Begins changing the query.one()/query.all() style to use explicit
select statements as part of the new style mandated for migration
to SQLAlchemy 2.0.
* Instead of using field label strings for joined queries, use the
object format, which makes much more sense now, and is part of
the items required for eventual migration to 2.0.
* DB queries involving Traits are now loaded using SelectInLoad
as opposed to Joins. The now deprecated ORM queries were quietly
and silently de-duplicating rows and providing consistent sets
from the resulting joined table responses, however putting much
higher CPU load on the processing of results on the client.
Prior performance testing has informed us this should be a minimal
overhead impact, however these queries should no longer be in
transactions with the Database Servers which should offset the
shift in load pattern. The reason we cannot continue to deduplicate
locally in our code is because we carry Dict data sets which cannot
be hashed for deduplication. Most projects have handled this by
treating them as Text and then converting, but without a massive
rewrite, this seems to be the viable middle ground.
* Adds an explict mapping for traits and tags on the Node object
to point directly to the NodeTrait and NodeTag classes. This
superceeds the prior usage of a backref to make the association.
* Splits SQLAlchemy class model Node into Node and NodeBase, which
allows for high performance queries to skip querying for ``tags``
and ``traits``. Otherwise with the afrormentioned lookups would
always execute as they are now properties as well on the Node
class. This more common of a SQLAlchemy model, but Ironic's model
has been a bit more rigid to date.
* Adds a ``start_consoles`` and ``start_allocations`` option to the
conductor ``init_host`` method. This allows unit tests to be
executed and launched with the service context, while *not* also
creating race conditions which resulted in failed tests.
* The db API ``_paginate_query`` wrapper now contains additional
logic to handle traditional ORM query responses and the newer style
of unified query responses. Due to differences in queries and handling,
which also was part of the driver for the creation of ``NodeBase``,
as SQLAlchemy will only create an object if a base object is referenced.
Also, by default, everything returned is a tuple in 1.4 with the
* Also modified one unit test which counted time.sleep calls, which is
a known pattern which can create failures which are ultimately noise.
Ultimately, I have labelled the remaining places which SQLAlchemy
warnings are raised at for deprecation/removal of functionality,
which needs to be addressed.
We will use this in the future to prepare for SQLAlchemy 2.0. For now,
we're simply using it to filter out some of the more annoying warnings
and to highlight general SQLAlchemy issues we need to address.
Signed-off-by: Stephen Finucane <firstname.lastname@example.org>
Create a new Sphinx extension called 'web_api_docstring' to process
docstrings from the API classes, in order to generate API
The lower-constraints test was removed becuase of an issue where pip
could not correctly determine the required packages versions to install,
ending in an almost infinite loop that would end up in timeout, failure,
and general mayhem.
Recently the issue has been fixed and, if properly configured, the
lower-constraints test can provide good indication of which minimum
versions are required to support the current code.
This patch adds the test back to the current development branch.
The long term goal is to keep the lower-constraints file in the stable
branches, but remove the test job to avoid issues during backports.
Adds policy scope based RBAC handling for the allocations
endpoing which enables admins to create allocations if
they have baremetal nodes which are available to them.
Usage of usedevelop=True results in all requirements.txt dependencies
always installed (without constraints) even when requirements.txt
is not explicitly added. Override usedevelop for these environments.
Add requirements.txt to the 'docs' environment since it imports
the whole ironic in the process.
* move pep8 dependencies from test-requirements to tox.ini,
they're not needed there and are hard to constraint properly.
* add oslo.cache to l-c to avoid bump of dependencies