Performance auditor fixture

[MrMarble]

Important

Please read the description before looking at the code, I've also added some comments to the code itself

I've created a test fixture that gathers some analytics to help analyze the performance of the application.
The idea is to not longer need django-silk or similar to conduct some manual tests to check/improve performance, just write a test case replicating the workflow or calling the function you want to test using this new fixture and review the results.

The fixture will print a summary of the results in the cli after the tests:

===================================================== Performance Regression Report =====================================================
-----------------------------------------------------------------------------------------------------------------------------------------
Test Name                                | Time     | CPU Time | DB Time  | Queries | Tables | N+1  | Writes | Dup Queries | Slow queries
-----------------------------------------------------------------------------------------------------------------------------------------
test_flaw_details_with_client            |    256ms |    235ms |     13ms |      58 |     15 |    3 |      0 |           1 |            0
test_flaw_details_with_factory           |     25ms |     23ms |      1ms |      10 |     12 |    0 |      0 |           0 |            0
test_list_endpoints[/flaws]              |    145ms |    141ms |      1ms |      14 |     13 |    0 |      0 |           0 |            0
test_list_endpoints[/flaws?include_hi... |    171ms |    160ms |      8ms |      14 |     23 |    0 |      0 |           0 |            0
test_list_endpoints[/flaws?exclude_fi... |    129ms |    123ms |      7ms |      13 |     13 |    0 |      0 |           0 |            0
test_list_endpoints[/flaws?include_fi... |     73ms |     65ms |      3ms |      24 |      2 |    2 |      0 |           0 |            0
test_list_endpoints[/affects]            |     62ms |     60ms |      3ms |       5 |      6 |    0 |      0 |           0 |            0
test_list_endpoints[/affects?include_... |     79ms |     74ms |      6ms |       6 |     11 |    0 |      0 |           0 |            0
test_fn_call                             |     57ms |     47ms |     15ms |      18 |      9 |    1 |      0 |           0 |            0
test_create_flaw                         |     33ms |     29ms |      1ms |      23 |      9 |    2 |      2 |           3 |            0

It will also generate a full report in markdown in to performance_report_YYYY-MM-DD_HH-MM.md so you can review old runs while iterating, this report will also be added to the github summary for the test action, you can check it here https://github.com/RedHatProductSecurity/osidb/actions/runs/19742723923?pr=1147.

The tests I wrote in this pull request are just an example of usage, not intended to be merged, we could keep the fixture alone for manual usage or ideally implement some kind of smoke tests that runs on release (detectable by branch name for example) and have these test generate the report, but we should discuss it.

I tried to keep it simple and objective without assuming or suggesting fixes/improvements that could be wrong, is just plain data to help understand the performance of the application.

[roduran-dev]

Very, nice change, with this we can introduce some bothersome cases like megaflaws and compare the performance.

2 questions:

• does this change the default behaviour on pytest output?
• this just changes only if performance_audit is used?

[MrMarble]

Very, nice change, with this we can introduce some bothersome cases like megaflaws and compare the performance.

2 questions:

* does this change the default behaviour on pytest output?

* this just changes only if performance_audit is used?

1. No it does not change the default behaviour of pytest
2. Yes, the report is only generated for tests using the performance_audit fixture.