Hv placement switch

[DavidFair]

• Drop support for getting VCPUs / Memory / Disk Usage from Hypervisor, as these are deprecated and return None
• Drop the enums to fail loud downstream as these will never work again
• Add placement query object with support for getting usage
• Major version bump to reflect API breaking changes to HypervisorQuery

TODO pending draft feedback:

• Docs
• Unit tests

Testing

from openstackquery import PlacementQuery

if __name__ == "__main__":
    query = PlacementQuery()
    query.select("name", "id")
    query.where("equal_to", "name", value="hv519.nubes.rl.ac.uk")
    query.run("dev")

    results = query.to_objects()
    print(results)

[codecov]

Codecov Report

Attention: Patch coverage is 69.09091% with 34 lines in your changes missing coverage. Please review.

Project coverage is 98.76%. Comparing base (ba9d83e) to head (7e80e57).

Files with missing lines	Patch %	Lines
openstackquery/runners/placement_runner.py	47.36%	20 Missing ⚠️
openstackquery/enums/props/placement_properties.py	73.07%	7 Missing ⚠️
openstackquery/mappings/placement_mapping.py	80.00%	6 Missing ⚠️
openstackquery/api/query_objects.py	66.66%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main       #9      +/-   ##
==========================================
- Coverage   99.55%   98.76%   -0.80%     
==========================================
  Files         111      115       +4     
  Lines        4244     4282      +38     
==========================================
+ Hits         4225     4229       +4     
- Misses         19       53      +34     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[DavidFair]

Questions for draft feedback:

• Any objections to the approach of getting the usage since the SDK doesn't support it (we can upstream this later), I've tried to match how upstream OpenStack SDK does it to make it easier to contribute
• Are we happy with serialising to our own dataclass, or will this cause problems?
• Should it be called Placement* or something else since we only really handle ResourceProviders from the Placement API?

[meoflynn]

• Should it be called Placement* or something else since we only really handle ResourceProviders from the Placement API?

I think it depends on whether we will add more features that interact with the Placement API in the future. If there are other features we want to add that will interact with Placement, then calling it Placement* would be okay. However if we are only going to be interacting with this specific part of the API we should call it something else.

[gmatthews20]

The switch to using the placement api for hypervisor resource usage was made in #1 which fixed the missing resource fields and was implemented so it was backwards compatible with the hypervisor query. After investigating the issue I think you were having with the find empty hypervisor stackstorm action there was a missing catch for KeyError in the client side filter handler when the resource is not found as is the case for ironic hypervisors.
I have fixed it here #10 and tested that it works with the find empty hypervisor action.