Remove Duckdb dependency (#19400)

* Update datadog_checks_base/CHANGELOG.md

Co-authored-by: Ilia Kurenkov <[email protected]>

* remove duckdb python

* dep

* readme

* readme

* remove duckdb

* remove duckdb

* remove duckdb

* commit

* commit

* commit

* remove duckdb

---------

Co-authored-by: Ilia Kurenkov <[email protected]>

LICENSE-3rdparty.csv

@@ -22,7 +22,6 @@ cryptography,PyPI,BSD-3-Clause,Copyright (c) Individual contributors.
 cryptography,PyPI,PSF,Copyright (c) Individual contributors.
 ddtrace,PyPI,BSD-3-Clause,"Copyright 2016 Datadog, Inc."
 dnspython,PyPI,ISC,Copyright (C) Dnspython Contributors
-duckdb,PyPI,MIT,Copyright (c) Hannes Muehleisen
 flup,Vendor,BSD-3-Clause,Copyright (c) 2005 Allan Saddi. All Rights Reserved.
 flup-py3,Vendor,BSD-3-Clause,"Copyright (c) 2005, 2006 Allan Saddi <[email protected]> All rights reserved."
 foundationdb,PyPI,Apache-2.0,Copyright 2017 FoundationDB

agent_requirements.in

@@ -13,7 +13,6 @@ confluent-kafka==2.8.0
 cryptography==43.0.1
 ddtrace==2.10.6
 dnspython==2.6.1
-duckdb==1.1.1
 foundationdb==6.3.24
 hazelcast-python-client==5.4.0
 in-toto==2.0.0

duckdb/README.md

@@ -27,6 +27,21 @@ Follow the instructions below to install and configure this check for an Agent r
 The DuckDB check is included in the [Datadog Agent][2] package.
 No additional installation is needed on your server.
 
+#### Dependencies
+
+The [duckdb][10] client library is required. To install it, ensure you have a working compiler and run:
+
+##### Unix
+
+```text
+sudo -Hu dd-agent /opt/datadog-agent/embedded/bin/pip install duckdb==1.1.1
+```
+##### Windows
+
+```text
+"C:\Program Files\Datadog\Datadog Agent\embedded3\python.exe" -m pip install duckdb==1.1.1
+```
+
 ### Configuration
 
 1. Edit the `duckdb.d/conf.yaml` file, in the `conf.d/` folder at the root of your Agent's configuration directory to start collecting your duckdb performance data. See the [sample duckdb.d/conf.yaml][4] for all available configuration options.
@@ -65,3 +80,4 @@ Need help? Contact [Datadog support][8].
 [7]: https://github.com/DataDog/integrations-core/blob/master/duckdb/metadata.csv
 [8]: https://docs.datadoghq.com/help/
 [9]: https://duckdb.org/docs/
+[10]: https://pypi.org/project/duckdb/

duckdb/datadog_checks/duckdb/check.py

@@ -8,7 +8,19 @@
 from contextlib import closing, contextmanager
 from copy import deepcopy
 
-import duckdb
+from datadog_checks.base.errors import CheckException
+
+try:
+    import duckdb
+
+    dk_import_error = None
+except ImportError as e:
+    duckdb = None
+    dk_import_error = e
+    raise CheckException(
+        "Duckdb was not imported correctly, make sure the library is installed."
+        "Please refer to datadog documentation for more details. Error is %s" % dk_import_error
+    )
 
 from datadog_checks.base import AgentCheck
 from datadog_checks.base.utils.db import QueryManager
@@ -46,21 +58,27 @@ def __init__(self, name, init_config, instances):
         self.check_initializations.append(self._query_manager.compile_queries)
 
     def check(self, _):
-        retry_delay = 5
-        max_retries = self.connection_attempt
-        for attempt in range(1, max_retries + 1):
-            try:
-                with self.connect() as conn:
-                    if conn:
-                        self._connection = conn
-                        self._query_manager.execute()
-                        break
-            except Exception as e:
-                self.log.warning('Unable to connect to the database:  "%s" , retrying...', e)
-                if attempt < max_retries:
-                    time.sleep(retry_delay)
-                else:
-                    self.log.error('Max connection retries reached')
+        if duckdb is None:
+            raise CheckException(
+                "Duckdb was not imported correctly, make sure the library is installed."
+                "Please refer to datadog documentation for more details. Error is %s" % dk_import_error
+            )
+        else:
+            retry_delay = 5
+            max_retries = self.connection_attempt
+            for attempt in range(1, max_retries + 1):
+                try:
+                    with self.connect() as conn:
+                        if conn:
+                            self._connection = conn
+                            self._query_manager.execute()
+                            break
+                except Exception as e:
+                    self.log.warning('Unable to connect to the database:  "%s" , retrying...', e)
+                    if attempt < max_retries:
+                        time.sleep(retry_delay)
+                    else:
+                        self.log.error('Max connection retries reached')
 
     def _execute_query_raw(self, query):
         with closing(self._connection.cursor()) as cursor:

duckdb/hatch.toml

@@ -1,5 +1,10 @@
 [env.collectors.datadog-checks]
 
+[envs.default]
+dependencies = [
+  "duckdb==1.1.1",
+]
+
 [[envs.default.matrix]]
 python = ["3.12"]
 version = ["1.1.1"]

duckdb/pyproject.toml

@@ -36,9 +36,7 @@ dynamic = [
 ]
 
 [project.optional-dependencies]
-deps = [
-    "duckdb==1.1.1",
-]
+deps = []
 
 [project.urls]
 Source = "https://github.com/DataDog/integrations-core"

duckdb/tests/conftest.py

@@ -7,10 +7,16 @@
 
 from . import common
 
+E2E_METADATA = {
+    'start_commands': [
+        'pip install duckdb==1.1.1',
+    ]
+}
+
 
 @pytest.fixture(scope='session')
 def dd_environment():
-    yield common.DEFAULT_INSTANCE
+    yield common.DEFAULT_INSTANCE, E2E_METADATA
 
 
 @pytest.fixture

duckdb/tests/test_e2e.py

@@ -5,6 +5,6 @@
 
 
 @pytest.mark.e2e
-def test_e2e(dd_agent_check):
-    aggregator = dd_agent_check()
+def test_e2e(dd_agent_check, instance):
+    aggregator = dd_agent_check(instance)
     aggregator.assert_all_metrics_covered()