This change adds configuration for the pre-commit tool,
follow-up changes will address the remaining issues in a phased
approach to make the reviews simpler.
This is based on the pre-commit config used in nova
with some additional hooks.
Follow-up changes will address the FIXME comments
related to sphinx-lint and codespell, as well as update tox
to enforce these checks in ci.
Change-Id: I87681a19f7fa88366c2b0d310c8b3153aa6a137b
The datetime.utcnow() is deprecated in Python 3.12.
Replace datetime.utcnow() with oslo_utils.timeutils.utcnow().
This bumps oslo.utils to 7.0.0.
Change-Id: Icccbb0549add686a744a72b354932471cbf91c92
Signed-off-by: Takashi Natsume <takanattie@gmail.com>
This code worked around a bug in eventlet[1] that has been fixed in
115103d5608cbe8f15df10e27eba1644f5364e95. The fix has been available in
every eventlet release since v0.27.0.
[1] https://github.com/eventlet/eventlet/issues/592
Co-Authored-By: Cyril Roelandt <cyril@redhat.com>
Change-Id: Ifc0b9c1d7f022db54c34c48c903a1719f9404d04
This was originally five patches, but they are all needed to pass
any of the test jobs now, so they have been squashed into one:
Co-Authored-By: Dan Smith (dms@danplanet.com)
First:
The autoload argument was removed[1] in SQLAlchemy and only
the autoload_with argument should be passed.
The autoload argument is set according to the autoload_with argument
automatically even in SQLAlchemy 1.x[2] so is not at all needed.
[1] c932123bac
[2] ad8f921e96
Second:
Remove _warn_on_bytestring for newer SA, AFAICT, this flag has been
removed from SQLAlchemy and that is why watcher-db-manage fails to
initialize the DB for me on jammy. This migration was passing the
default value (=False) anyway, so I assume this is the right "fix".
Third:
Fix joinedload passing string attribute names
Fourth:
Fix engine.select pattern to use begin() per the migration guide.
Fifth:
Override the apscheduler get_next_run_time() which appears to be
trivially not compatible with SQLAlchemy 2.0 because of a return type
from scalar().
Change-Id: I000e5e78f97f82ed4ea64d42f1c38354c3252e08
Minimal refactor of SQLAlchemy api module to be compatible with
oslo.db >= 15.0.0 where autocommit behaviour was dropped.
Closes-Bug: #2056181
Change-Id: I33be53f647faae2aad30a43c10980df950d5d7c2
At the moment, Watcher can use a single bare metal provisioning
service: Openstack Ironic.
We're now adding support for Canonical's MAAS service [1], which
is commonly used along with Juju [2] to deploy Openstack.
In order to do so, we're building a metal client abstraction, with
concrete implementations for Ironic and MAAS. We'll pick the MAAS
client if the MAAS url is provided, otherwise defaulting to Ironic.
For now, we aren't updating the baremetal model collector since it
doesn't seem to be used by any of the existing Watcher strategy
implementations.
[1] https://maas.io/docs
[2] https://juju.is/docs
Implements: blueprint maas-support
Change-Id: I6861995598f6c542fa9c006131f10203f358e0a6
Power-off actions created by the energy saving strategy include
a resource name property, which currently isn't part of the
action json schema. For this reason, json schema validation fails.
Additional properties are not allowed ('resource_name' was unexpected)
We'll update the json schema, including the resource name property.
Change-Id: I924d36732a917c0be98b08c2f4128e9136356215
A couple of object tests are failing, probably after a dependency
bump.
watcher.objects.base.objects is mocked, so the registered object
version isn't properly retrieved, leading to a type error:
File "/mnt/data/workspace/watcher/watcher/tests/objects/test_objects.py",
line 535, in test_hook_chooses_newer_properly
reg.registration_hook(MyObj, 0)
File "/mnt/data/workspace/watcher/watcher/objects/base.py",
line 46, in registration_hook
cur_version = versionutils.convert_version_to_tuple(
File "/home/ubuntu/openstack_venv/lib/python3.10/site-packages/oslo_utils/versionutils.py",
line 91, in convert_version_to_tuple
version_str = re.sub(r'(\d+)(a|alpha|b|beta|rc)\d+$', '\\1', version_str)
File "/usr/lib/python3.10/re.py", line 209, in sub
return _compile(pattern, flags).sub(repl, string, count)
TypeError: expected string or bytes-like object
We'll solve the issue by setting the VERSION attribute against
the mock object.
Change-Id: Ifeb38b98f1d702908531de5fc5c846bd1c53de4b
The "vm workload consolidation" strategy is summing up instance
usage in order to estimate host usage.
The problem is that some infrastructure services (e.g. OVS or Ceph
clients) may also use a significant amount of resources, which
would be ignored. This can impact Watcher's ability to detect
overloaded nodes and correctly rebalance the workload.
This commit will use the host metrics, if available. The proposed
implementation uses the maximum value between the host metric
and the sum of the instance metrics.
Note that we're holding a dict of host metric deltas in order to
account for planned migrations.
Change-Id: I82f474ee613f6c9a7c0a9d24a05cba41d2f68edb
The "cpu_util" metric has been deprecated a few years ago.
We'll obtain the same result by converting the cumulative cpu
time to a percentage, leveraging the rate of change aggregation.
Change-Id: I18fe0de6f74c785e674faceea0c48f44055818fe
There may be no available metrics for instances that are stopped
or were recently spawned. This makes retries unnecessary and time
consuming.
For this reason, we'll ignore gnocchi MetricNotFound errors.
Change-Id: I79cd03bf04db634b931d6dfd32d5150f58e82044
We're adding a few info log messages in order to trace the
"vm consolidation" strategy more easily.
Change-Id: I8ce1a9dd173733f1b801839d3ad0c1269c4306bb
Although Watcher supports cold migrations, the vm workload
consolidation workflow only allows live migrations to be
performed.
We'll remove this unnecessary limitation so that stopped instances
could be cold migrated.
Change-Id: I4b41550f2255560febf8586722a0e02045c3a486
The Nova collector json schema validation started [1][2] failing after
the jsonschema upper constraint was bumped from 4.17.3 to 4.19.1 [3].
The reason is that jsonschema v4.18.0a1 switched to a reference
resolving library [4], which treats the aggregate "id" as a jsonschema
id and expects it to be a string [5]. For this reason, we're now getting
AttributeError exceptions.
As a workaround, we'll rename the "id" ref element as "host_aggr_id".
Also, the watcher-tempest-multinode job is configured to use Focal,
which is no longer supported by Devstack [6]. That being considered,
we'll switch to Ubuntu Jammy (22.04).
While at it, we're disabling Cinder Backup, which isn't used while
testing Watched. It currently causes Devstack failures since it
uses the Swift backend by default, which is disabled.
[1] https://paste.opendev.org/raw/bjQ1uIdbDMnmA1UEhxLL/
[2] https://paste.opendev.org/raw/bNgxqulBwBLYB7tNhrU4/
[3] ab0dcbdda2
[4] https://github.com/python-jsonschema/jsonschema/releases/tag/v4.18.0a1
[5] c23a5dc1c9/referencing/jsonschema.py (L54-L55C18)
[6] https://paste.openstack.org/raw/bSoSyXgbtmq6d9768HQn/
Change-Id: I300620c2ec4857b1e0d402a9b57a637f576eeb24
oslo.db 12.1.0 has changed the default value for the 'autocommit'
parameter of 'LegacyEngineFacade' from 'True' to 'False'. This is a
necessary step to ensure compatibility with SQLAlchemy 2.0. However, we
are currently relying on the autocommit behavior and need changes to
explicitly manage sessions. Until that happens, we need to override the
default.
Co-Authored-By: Stephen Finucane <stephenfin@redhat.com>
Change-Id: I7db39d958d087322bfa0aad70dfbd04de9228dd7
>>> random.sample([5,10], 1.3)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib64/python3.6/random.py", line 321, in sample
result = [None] * k
TypeError: can't multiply sequence by non-int of type 'float'
Change-Id: Ifa5dca06f07220512579e4fe3c5c741aeffc71cc
Block Storage API v2 was deprecated during Pike cycle and is being
removed during Xena cycle, and current v3 API should be used instead.
Change-Id: Ia5247742b31f5f07186ef908588f0972d3ac609f
Python introduced http.HTTPStatus since version 3.5,
and Wallaby has targeted a minimum version of python 3.6.
Change-Id: I45f732f0f59b8fae831bb6c07f4fdd98cdd7409a
The Query.with_lockmode() method is deprecated since version 0.9.0
and will be removed in a future release. [1]
This patch replaces it with Query.with_for_update().
The 'faultstring' was been modified to 'Exactly 5 or 6 columns has to be
specified for iterator expression', so adds one space between "iterator"
and "expression" for 'expected_error_msg'.
Also use upper-constraints in doc build to avoid issues in pdf build.
[1]
https://docs.sqlalchemy.org/en/13/orm/query.html#sqlalchemy.orm.query.Query.with_lockmode
Closes-Bug: #1933226
Change-Id: I0ad514da647bb08790259fd27e56a41f6dbbbaa0
As per the community goal of migrating the policy file
the format from JSON to YAML[1], we need to do two things:
1. Change the default value of '[oslo_policy] policy_file''
config option from 'policy.json' to 'policy.yaml' with
upgrade checks.
2. Deprecate the JSON formatted policy file on the project side
via warning in doc and releasenotes.
Also replace policy.json to policy.yaml ref from doc and tests.
[1]https://governance.openstack.org/tc/goals/selected/wallaby/migrate-policy-format-from-json-to-yaml.html
Change-Id: I207c02ba71fe60635fd3406c9c9364c11f259bae
There is a commonly shared and proven rpc pattern used
across most OpenStack services that is already implemented
in watcher, but the functions are not used.
This patch basically makes use of the existing
rpc classes and removes some unnecessary code.
Change-Id: I57424561e0675a836d10b712ef1579a334f72018
Implements base method as well as some basic implementations to
retrieve time series metrics. Ceilometer can not be supported
as API documentation has been unavailable. Grafana will be
supported in follow-up patch.
Partially Implements: blueprint time-series-framework
Change-Id: I55414093324c8cff379b28f5b855f41a9265c2d3
