Exclude Scarf Usage Data Collection in CI Environments (#44155)

Most of the the CI systems add "CI=true" env var.

Refereces:

- https://docs.pytest.org/en/stable/explanation/ci.html
- https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/store-information-in-variables#default-environment-variables
- https://docs.travis-ci.com/user/environment-variables/
- https://docs.gitlab.com/ee/ci/variables/predefined_variables.html
- https://circleci.com/docs/variables/#built-in-environment-variables
- https://www.jenkins.io/doc/book/pipeline/jenkinsfile/#using-environment-variables
- https://adamj.eu/tech/2020/03/09/detect-if-your-tests-are-running-on-ci/
- https://github.com/The-Compiler/pytest-vw/blob/master/pytest_vw.py

Author: kaxil

airflow/utils/usage_data_collection.py

@@ -25,6 +25,7 @@
 
 from __future__ import annotations
 
+import os
 import platform
 from urllib.parse import urlencode
 
@@ -43,6 +44,10 @@ def usage_data_collection():
     if _version_is_prerelease(airflow_version):
         return
 
+    # Exclude CI environments
+    if _is_ci_environ():
+        return
+
     scarf_domain = "https://apacheairflow.gateway.scarf.sh/scheduler"
 
     try:
@@ -70,6 +75,26 @@ def _version_is_prerelease(version: str) -> bool:
     return parse(version).is_prerelease
 
 
+def _is_ci_environ() -> bool:
+    """Return True if running in any known CI environment."""
+    if os.getenv("CI") == "true":
+        # Generic CI variable set by many CI systems (GH Actions, Travis, GitLab, CircleCI, Jenkins, Heroku)
+        return True
+
+    # Other CI variables set by specific CI systems
+    ci_env_vars = {
+        "CIRCLECI",  # CircleCI
+        "CODEBUILD_BUILD_ID",  # AWS CodeBuild
+        "GITHUB_ACTIONS",  # GitHub Actions
+        "GITLAB_CI",  # GitLab CI
+        "JENKINS_URL",  # Jenkins
+        "TF_BUILD",  # Azure Pipelines
+        "TRAVIS",  # Travis CI
+    }
+
+    return any(var in os.environ for var in ci_env_vars)
+
+
 def get_platform_info() -> tuple[str, str]:
     return platform.system(), platform.machine()
 

tests/utils/test_usage_data_collection.py

@@ -43,12 +43,14 @@ def test_scarf_analytics_disabled(mock_get, is_enabled, is_prerelease):
 
 @mock.patch("airflow.settings.is_usage_data_collection_enabled", return_value=True)
 @mock.patch("airflow.utils.usage_data_collection._version_is_prerelease", return_value=False)
+@mock.patch("airflow.utils.usage_data_collection._is_ci_environ", return_value=False)
 @mock.patch("airflow.utils.usage_data_collection.get_database_version", return_value="12.3")
 @mock.patch("airflow.utils.usage_data_collection.get_database_name", return_value="postgres")
 @mock.patch("httpx.get")
 def test_scarf_analytics(
     mock_get,
     mock_is_usage_data_collection_enabled,
+    mock_version_is_ci,
     mock_version_is_prerelease,
     get_database_version,
     get_database_name,