[MAINTENANCE] Ajout de l'évaluation de performance des routes dans GeoNature

[jacquesfize]

This PR includes perfomance evaluation of specific routes in GeoNature backend. We use pytest-benchmark module to compute and store execution time of a designated python function. To automate the generation of pytest function to run a benchmark on a route, we designed a class name BenchmarkTest.

TODO LIST:

• Complete the declaration of benchmarked routes
• Build a PostgreSQL dump of an instance of GeoNature with a large amount of data
• Find a way to store and reuse execution time for comparison in a CI
• Compare results from different runs on the same commit

Flask routes included in the benchmark

Module	Route
Metadata	gn_meta.get_acquisition_frameworks_list
Metadata	gn_meta.get_datasets
OccHab	occhab.get_station
OccHab	occhab.list_stations (with or without restricted permissions)
OccHab	occhab.export_all_habitats (with or without restricted permissions)
Occtax	pr_occtax.getReleves
Synthese	gn_synthese.getDefaultsNomenclatures
Synthese	gn_synthese.get_observations_for_web (polygon, bbox, taxon_tree, departement, region, city) (with/without blurring)
Synthese	gn_synthese.export_observations_web (10^3,10^4,10^5,10^6 obs.) (with/without blurring)
Synthese	gn_synthese.export_status (10^3,10^4,10^5,10^6 obs.) (with/without blurring)
Synthese	gn_synthese.export_taxon_web (10^3,10^4,10^5,10^6 obs.) (with/without blurring)

First results

Synthese routes results

Occhab routes resuts

Occtax routes results

How to compare with previous benchmark

pytest --benchmark-only --benchmark-compare-fail="mean:0.00001" --benchmark-compare=<fichier_vieux_benchmark.json>

• --benchmark-compare-fail allows to indicate the condition of non-failure
• --benchmark-compare is the JSON files storing the last stable performances

To store the JSON containing the last accepted performance, we can use dedicated online storage like http://geonature.fr/data.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 78.53%. Comparing base (2a0582d) to head (f5d7190).

❗ Current head f5d7190 differs from pull request most recent head e77cd83. Consider uploading reports for the commit e77cd83 to get more accurate results

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #2896   +/-   ##
========================================
  Coverage    78.53%   78.53%           
========================================
  Files           89       89           
  Lines         7236     7236           
========================================
  Hits          5683     5683           
  Misses        1553     1553           

Flag	Coverage Δ
pytest	78.53% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

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