Add plugin data to scarf usage data collection

airflow/utils/usage_data_collection.py

@@ -33,6 +33,7 @@
 
 from airflow import __version__ as airflow_version, settings
 from airflow.configuration import conf
+from airflow.plugins_manager import get_plugin_info
 
 
 def usage_data_collection():
@@ -95,3 +96,15 @@ def get_executor() -> str:
 
 def get_python_version() -> str:
     return platform.python_version()
+
+
+def get_plugin_counts() -> dict[str, int]:
+    plugin_info = get_plugin_info()
+
+    return {
+        "plugins": len(plugin_info),
+        "flask_blueprints": sum(len(x["flask_blueprints"]) for x in plugin_info),
+        "appbuilder_views": sum(len(x["appbuilder_views"]) for x in plugin_info),
+        "appbuilder_menu_items": sum(len(x["appbuilder_menu_items"]) for x in plugin_info),
+        "timetables": sum(len(x["timetables"]) for x in plugin_info),
+    }

airflow/www/views.py

@@ -234,15 +234,22 @@ def build_scarf_url(dags_count: int) -> str:
     db_name = usage_data_collection.get_database_name()
     executor = usage_data_collection.get_executor()
     python_version = usage_data_collection.get_python_version()
+    plugin_counts = usage_data_collection.get_plugin_counts()
+    plugins_count = plugin_counts["plugins"]
+    flask_blueprints_count = plugin_counts["flask_blueprints"]
+    appbuilder_views_count = plugin_counts["appbuilder_views"]
+    appbuilder_menu_items_count = plugin_counts["appbuilder_menu_items"]
+    timetables_count = plugin_counts["timetables"]
 
     # Path Format:
-    # /{version}/{python_version}/{platform}/{arch}/{database}/{db_version}/{executor}/{num_dags}
+    # /{version}/{python_version}/{platform}/{arch}/{database}/{db_version}/{executor}/{num_dags}/{plugin_count}/{flask_blueprint_count}/{appbuilder_view_count}/{appbuilder_menu_item_count}/{timetables}
     #
     # This path redirects to a Pixel tracking URL
     scarf_url = (
         f"{scarf_domain}/webserver"
         f"/{version}/{python_version}"
         f"/{platform_sys}/{platform_arch}/{db_name}/{db_version}/{executor}/{dags_count}"
+        f"/{plugins_count}/{flask_blueprints_count}/{appbuilder_views_count}/{appbuilder_menu_items_count}/{timetables_count}"
     )
 
     return scarf_url

docs/apache-airflow/faq.rst

@@ -546,3 +546,5 @@ The telemetry data collected is limited to the following:
 - Executor
 - Metadata DB type & its version
 - Number of DAGs
+- Number of Airflow plugins
+- Number of timetables, Flask blueprints, Flask AppBuilder views, and Flask Appbuilder menu items from Airflow plugins

tests/www/views/test_views.py

@@ -599,26 +599,35 @@ def test_invalid_dates(app, admin_client, url, content):
     assert re.search(content, resp.get_data().decode())
 
 
-@pytest.mark.parametrize("enabled, dags_count", [(False, 5), (True, 5)])
+@pytest.mark.parametrize("enabled", [False, True])
 @patch("airflow.utils.usage_data_collection.get_platform_info", return_value=("Linux", "x86_64"))
 @patch("airflow.utils.usage_data_collection.get_database_version", return_value="12.3")
 @patch("airflow.utils.usage_data_collection.get_database_name", return_value="postgres")
 @patch("airflow.utils.usage_data_collection.get_executor", return_value="SequentialExecutor")
 @patch("airflow.utils.usage_data_collection.get_python_version", return_value="3.8.5")
+@patch("airflow.utils.usage_data_collection.get_plugin_counts")
 def test_build_scarf_url(
-    get_platform_info,
-    get_database_version,
-    get_database_name,
-    get_executor,
+    get_plugin_counts,
     get_python_version,
+    get_executor,
+    get_database_name,
+    get_database_version,
+    get_platform_info,
     enabled,
-    dags_count,
 ):
+    get_plugin_counts.return_value = {
+        "plugins": 10,
+        "flask_blueprints": 15,
+        "appbuilder_views": 20,
+        "appbuilder_menu_items": 25,
+        "timetables": 30,
+    }
     with patch("airflow.settings.is_usage_data_collection_enabled", return_value=enabled):
-        result = build_scarf_url(dags_count)
+        result = build_scarf_url(5)
         expected_url = (
             "https://apacheairflow.gateway.scarf.sh/webserver/"
             f"{airflow_version}/3.8.5/Linux/x86_64/postgres/12.3/SequentialExecutor/5"
+            f"/10/15/20/25/30"
         )
         if enabled:
             assert result == expected_url