Remove Scarf tracking (apache#45865)

astronomer · Jan 22, 2025 · 70a8a8a · 70a8a8a
1 parent 6751c96
commit 70a8a8a
Show file tree

Hide file tree

Showing 11 changed files with 2 additions and 318 deletions.
diff --git a/README.md b/README.md
@@ -536,6 +536,3 @@ The CI infrastructure for Apache Airflow has been sponsored by:
 
 <a href="https://astronomer.io"><img src="https://assets2.astronomer.io/logos/logoForLIGHTbackground.png" alt="astronomer.io" width="250px"></a>
 <a href="https://aws.amazon.com/opensource/"><img src="docs/integration-logos/aws/[email protected]" alt="AWS OpenSource" width="130px"></a>
-
-<!-- telemetry/analytics pixel: -->
-<img referrerpolicy="no-referrer-when-downgrade" src="https://static.scarf.sh/a.png?x-pxid=1b5a5e3c-da81-42f5-befa-42d836bf1b54" alt="Tracking Pixel" />
diff --git a/RELEASE_NOTES.rst b/RELEASE_NOTES.rst
@@ -220,7 +220,6 @@ Scarf based telemetry: Airflow now collect telemetry data (#39510)
 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
 Airflow integrates Scarf to collect basic usage data during operation. Deployments can opt-out of data collection by
 setting the ``[usage_data_collection]enabled`` option to ``False``, or the ``SCARF_ANALYTICS=false`` environment variable.
-See :ref:`Usage data collection FAQ <usage-data-collection>` for more information.
 
 Datasets no longer trigger inactive DAGs (#38891)
 """""""""""""""""""""""""""""""""""""""""""""""""

diff --git a/airflow/cli/commands/local_commands/scheduler_command.py b/airflow/cli/commands/local_commands/scheduler_command.py
@@ -32,7 +32,6 @@
 from airflow.utils import cli as cli_utils
 from airflow.utils.providers_configuration_loader import providers_configuration_loaded
 from airflow.utils.scheduler_health import serve_health_check
-from airflow.utils.usage_data_collection import usage_data_collection
 
 log = logging.getLogger(__name__)
 
@@ -50,8 +49,6 @@ def scheduler(args: Namespace):
     """Start Airflow Scheduler."""
     print(settings.HEADER)
 
-    usage_data_collection()
-
     run_command_with_daemon_option(
         args=args,
         process_name="scheduler",

diff --git a/airflow/config_templates/config.yml b/airflow/config_templates/config.yml
@@ -2625,28 +2625,6 @@ sensors:
       type: float
       example: ~
       default: "604800"
-usage_data_collection:
-  description: |
-    Airflow integrates `Scarf <https://about.scarf.sh/>`__ to collect basic platform and usage data
-    during operation. This data assists Airflow maintainers in better understanding how Airflow is used.
-    Insights gained from this telemetry are critical for prioritizing patches, minor releases, and
-    security fixes. Additionally, this information supports key decisions related to the development road map.
-    Check the FAQ doc for more information on what data is collected.
-
-    Deployments can opt-out of analytics by setting the ``enabled`` option
-    to ``False``, or the ``SCARF_ANALYTICS=false`` environment variable.
-    Individual users can easily opt-out of analytics in various ways documented in the
-    `Scarf Do Not Track docs <https://docs.scarf.sh/gateway/#do-not-track>`__.
-
-  options:
-    enabled:
-      description: |
-        Enable or disable usage data collection and sending.
-      version_added: 2.10.0
-      type: boolean
-      example: ~
-      default: "True"
-      see_also: ":ref:`Usage data collection FAQ <usage-data-collection>`"
 dag_bundles:
   description: |
     Configuration for the DAG bundles. This allows Airflow to load DAGs from different sources.

diff --git a/airflow/reproducible_build.yaml b/airflow/reproducible_build.yaml
@@ -1,2 +1,2 @@
-release-notes-hash: f1d91d32ade6da6eedd24362610d5f84
-source-date-epoch: 1734354109
+release-notes-hash: ab7a935709e7a13d5587b7eb727ae2bd
+source-date-epoch: 1737531923
diff --git a/airflow/settings.py b/airflow/settings.py
@@ -652,13 +652,6 @@ def initialize():
     atexit.register(dispose_orm)
 
 
-def is_usage_data_collection_enabled() -> bool:
-    """Check if data collection is enabled."""
-    return conf.getboolean("usage_data_collection", "enabled", fallback=True) and (
-        os.getenv("SCARF_ANALYTICS", "").strip().lower() != "false"
-    )
-
-
 # Const stuff
 
 KILOBYTE = 1024

diff --git a/airflow/utils/usage_data_collection.py b/airflow/utils/usage_data_collection.py
diff --git a/docs/apache-airflow/faq.rst b/docs/apache-airflow/faq.rst
@@ -519,26 +519,3 @@ This means ``explicit_defaults_for_timestamp`` is disabled in your mysql server
 
 #. Set ``explicit_defaults_for_timestamp = 1`` under the ``mysqld`` section in your ``my.cnf`` file.
 #. Restart the Mysql server.
-
-Does Airflow collect any telemetry data?
-----------------------------------------
-
-.. _usage-data-collection:
-
-Airflow integrates `Scarf <https://about.scarf.sh/>`__ to collect basic usage data during operation.
-This data assists Airflow maintainers in better understanding how Airflow is used.
-Insights gained from this data are helpful for prioritizing patches, minor releases, and
-security fixes. Additionally, this information supports key decisions related to the development road map.
-
-Deployments can opt-out of data collection by setting the :ref:`[usage_data_collection] enabled <config:usage_data_collection__enabled>`
-option to ``False``, or the ``SCARF_ANALYTICS=false`` environment variable.
-Individual users can easily opt-out of analytics in various ways documented in the
-`Scarf Do Not Track docs <https://docs.scarf.sh/gateway/#do-not-track>`__.
-
-The telemetry data collected is limited to the following:
-
-- Airflow version
-- Python version
-- Operating system & machine architecture
-- Executor
-- Metadata DB type & its version
diff --git a/docs/apache-airflow/installation/installing-from-pypi.rst b/docs/apache-airflow/installation/installing-from-pypi.rst
@@ -330,12 +330,6 @@ dependencies compatible with just airflow core at the moment Airflow was release
     # For example: https://raw.githubusercontent.com/apache/airflow/constraints-|version|/constraints-no-providers-3.9.txt
     pip install "apache-airflow==${AIRFLOW_VERSION}" --constraint "${CONSTRAINT_URL}"
 
-
-.. note::
-
-    Airflow uses `Scarf <https://about.scarf.sh/>`__ to collect basic usage data during operation.
-    Check the :ref:`Usage data collection FAQ <usage-data-collection>` for more information about the data collected and how to opt-out.
-
 Troubleshooting
 '''''''''''''''
 

diff --git a/tests/core/test_settings.py b/tests/core/test_settings.py
@@ -27,7 +27,6 @@
 import pytest
 
 from airflow.exceptions import AirflowClusterPolicyViolation, AirflowConfigException
-from airflow.settings import is_usage_data_collection_enabled
 
 from tests_common.test_utils.config import conf_vars
 
@@ -294,26 +293,3 @@ def test_encoding_absent_in_v2(is_v1, mock_conf):
         engine_args = settings.prepare_engine_args()
 
         assert "encoding" not in engine_args
-
-
-@pytest.mark.parametrize(
-    "env_var, conf_setting, is_enabled",
-    [
-        ("false", "True", False),  # env forces disable
-        ("false", "False", False),  # Both force disable
-        ("False ", "False", False),  # Both force disable
-        ("true", "True", True),  # Both enable
-        ("true", "False", False),  # Conf forces disable
-        (None, "True", True),  # Default env, conf enables
-        (None, "False", False),  # Default env, conf disables
-    ],
-)
-def test_usage_data_collection_disabled(env_var, conf_setting, is_enabled):
-    conf_patch = conf_vars({("usage_data_collection", "enabled"): conf_setting})
-
-    if env_var is not None:
-        with conf_patch, patch.dict(os.environ, {"SCARF_ANALYTICS": env_var}):
-            assert is_usage_data_collection_enabled() == is_enabled
-    else:
-        with conf_patch:
-            assert is_usage_data_collection_enabled() == is_enabled
diff --git a/tests/utils/test_usage_data_collection.py b/tests/utils/test_usage_data_collection.py