Tech Debt: Upgrading boto to boto3 for api & ingester. Needed to use newer moto needed by client.

Author: huzaifams-svg

.github/workflows/actions.yml

@@ -3,14 +3,15 @@ on: [push, pull_request]
 
 jobs:
   test-client:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-latest
+    timeout-minutes: 30  # Add this line
     strategy:
       matrix:
         python: [3.8, 3.9, "3.10", 3.12]
         extras: ["test", "test,queuable,sentry"]
     steps:
       - name: Setup Python
-        uses: actions/setup-python@v2.2.2
+        uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python }}
       - name: Check out repository code
@@ -20,13 +21,16 @@ jobs:
       - name: Test
         working-directory: ./client
         run: |
-          pip install -e .[${{ matrix.extras }}]
-          py.test
+          pip --version
+          pip install --verbose .[${{ matrix.extras }}]
+
+          # Run pytest with specific path and import mode
+          python -m pytest --import-mode=importlib ./datalake/tests/
   test-docker:
     runs-on: ubuntu-latest
     steps:
       - name: Check out repository code
-        uses: actions/checkout@v2
+        uses: actions/checkout@v4
         with:
           fetch-depth: 0
       - name: Test

.gitignore

@@ -58,4 +58,6 @@ target/
 
 # build artifacts
 version.txt
-.claude/settings.local.json
+
+# Claude Code
+.claude/

Dockerfile

@@ -7,29 +7,34 @@ ENV	LC_ALL C.UTF-8
 
 # TODO: keep requirements in one place
 RUN pip install \
-    blinker>=1.4 \
-    boto3>=1.1.3 \
-    click>=5.1 \
-    Flask>=0.10.1 \
-    flask-swagger>=0.2.14 \
-    memoized_property>=1.0.1 \
-    python-dateutil>=2.4.2 \
-    python-dotenv>=0.1.3 \
-    pytz>=2015.4 \
-    sentry-sdk[flask]>=0.19.5 \
-    requests>=2.5 \
-    simplejson>=3.3.1 \
-    six>=1.10.0 \
-    # test requirements
+    'boto==2.49.0' \
+    'boto3==1.35.41' \
+    'botocore==1.35.64' \
+    'datalake<2' \
     'flake8>=2.5.0,<4.1' \
     'freezegun<1' \
-    'moto<3' \
+    'moto>=5,<6' \
     'pytest<8' \
+    'pytest-xdist' \
     'responses<0.22.0' \
-    pyinotify>=0.9.4, \
-    raven>=5.0.0 \
     'tox>4,<5' \
-    'datalake<2'
+    # test requirements
+    'pytest-cov>=2.5.1,<4' \
+    'blinker>=1.4' \
+    'click>=5.1' \
+    'flask-swagger>=0.2.14' \
+    'Flask>=0.10.1' \
+    'memoized_property>=1.0.1' \
+    'pyinotify>=0.9.4' \
+    python-dateutil>=2.4.2 \
+    python-dotenv>=0.1.3 \
+    pytz>=2015.4 \
+    raven>=5.0.0 \
+    requests>=2.5 \
+    sentry-sdk[flask]>=0.19.5 \
+    simplejson>=3.3.1 \
+    six>=1.10.0
+
 
 RUN mkdir -p /opt/
 COPY . /opt/

api/tests/conftest.py

@@ -18,7 +18,7 @@
     ClientError as BotoClientError,
     NoCredentialsError
 )
-from moto import mock_dynamodb2
+from moto import mock_aws
 
 from datalake_api import app as datalake_api
 from datalake.tests import *  # noqa
@@ -71,7 +71,7 @@ def client():
 
 @pytest.fixture
 def dynamodb(request):
-    mock = mock_dynamodb2()
+    mock = mock_aws()
     mock.start()
 
     def tear_down():

client/datalake/tests/conftest.py

@@ -19,15 +19,10 @@
 import six
 
 
-try:
-    from moto import mock_s3
-    import boto3
-    from six.moves.urllib.parse import urlparse
-    import json
-except ImportError:
-    # if developers use s3-test features without having installed s3 stuff,
-    # things will fail. So it goes.
-    pass
+from moto import mock_aws
+import boto3
+from six.moves.urllib.parse import urlparse
+import json
 
 
 @pytest.fixture
@@ -142,7 +137,7 @@ def tear_down():
 
 @pytest.fixture
 def s3_connection(aws_connector):
-    with mock_s3():
+    with mock_aws():
         yield boto3.resource('s3')
 
 

client/pyproject.toml

@@ -37,7 +37,7 @@ dynamic = ["version"]
 test = [
     'pytest<8.0.0',
     'pytest-cov>=2.5.1,<4',
-    'moto[s3]>4,<5',
+    'moto[s3]>5,<6',
     'twine<4.0.0',
     'pip>=20.0.0,<22.0.0',
     'wheel<0.38.0',
@@ -73,7 +73,7 @@ distance-dirty = "{base_version}+{distance}.{vcs}{rev}.dirty"
 # Example formatted version: 1.2.3+42.ge174a1f.dirty
 
 [tool.pytest.ini_options]
-addopts = "--cov=planet.mc_client --cov-config .coveragerc"
+addopts = "--cov=datalake --cov-config .coveragerc"
 markers = [
   "slow: marks tests as slow (deselect with '-m \"not slow\"')"
 ]

ingester/datalake_ingester/queue.py

@@ -13,7 +13,8 @@
 # the License.
 
 from memoized_property import memoized_property
-import boto.sqs
+import boto3
+from botocore.exceptions import ClientError
 import simplejson as json
 import logging
 import os
@@ -39,16 +40,16 @@ def set_handler(self, h):
         self.handler = h
 
     @memoized_property
-    def _queue(self):
-        return self._connection.get_queue(self.queue_name)
+    def _connection(self):
+        # boto2 defaulted to us-east-1
+        region = os.environ.get('AWS_REGION', 'us-east-1')
+        return boto3.resource('sqs', region_name=region)
 
     @memoized_property
-    def _connection(self):
-        region = os.environ.get('AWS_REGION')
-        if region:
-            return boto.sqs.connect_to_region(region)
-        else:
-            return boto.connect_sqs()
+    def _queue(self):
+        # Let boto3 raise ClientError naturally if queue doesn't exist
+        # The error will include the queue name and a clear error code
+        return self._connection.get_queue_by_name(QueueName=self.queue_name)
 
     _LONG_POLL_TIMEOUT = 20
 
@@ -57,23 +58,38 @@ def drain(self, timeout=None):
         '''
         long_poll_timeout = timeout or self._LONG_POLL_TIMEOUT
         while True:
-            raw_msg = self._queue.read(wait_time_seconds=long_poll_timeout)
-            if raw_msg is None:
+            messages = self._queue.receive_messages(
+                MaxNumberOfMessages=1,
+                WaitTimeSeconds=long_poll_timeout
+            )
+            if not messages:
                 if timeout:
                     return
                 else:
                     continue
-            self._handle_raw_message(raw_msg)
+            self._handle_raw_message(messages[0])
 
     def _handle_raw_message(self, raw_msg):
         # eliminate newlines in raw message so it all logs to one line
-        raw = raw_msg.get_body().replace('\n', ' ')
+        raw = raw_msg.body.replace('\n', ' ')
         if not self.handler:
             self.logger.error('NO HANDLER CONFIGURED: %s', raw)
             return
 
         self.logger.info('RECEIVED: %s', raw)
-        msg = json.loads(raw)
 
+        try:
+            msg = json.loads(raw)
+        except Exception as e:
+            self.logger.warning('Failed to parse JSON (%s), using raw message: %s', str(e), raw)
+            msg = raw
+
+        # Process the message with the handler
+        # If this fails, the message stays in the queue and will be retried
+        # after the visibility timeout expires
         self.handler(msg)
-        self._queue.delete_message(raw_msg)
+
+        # Only delete the message after successful processing
+        # This ensures message reliability - if processing fails above,
+        # the message remains in the queue for retry
+        raw_msg.delete()

ingester/datalake_ingester/reporter.py

@@ -1,4 +1,4 @@
-import boto.sns
+import boto3
 import simplejson as json
 import logging
 from memoized_property import memoized_property
@@ -25,13 +25,11 @@ def _log_name(self):
 
     @memoized_property
     def _connection(self):
-        region = os.environ.get('AWS_REGION')
-        if region:
-            return boto.sns.connect_to_region(region)
-        else:
-            return boto.connect_sns()
+        # boto2 defaulted to us-east-1
+        region = os.environ.get('AWS_REGION', 'us-east-1')
+        return boto3.client('sns', region_name=region)
 
     def report(self, ingestion_report):
         message = json.dumps(ingestion_report)
         self.logger.info('REPORTING: %s', message)
-        self._connection.publish(topic=self.report_key, message=message)
+        self._connection.publish(TopicArn=self.report_key, Message=message)

ingester/datalake_ingester/storage.py

@@ -14,9 +14,8 @@
 
 
 from memoized_property import memoized_property
-import boto.dynamodb2
-from boto.dynamodb2.table import Table
-from boto.dynamodb2.exceptions import ConditionalCheckFailedException
+import boto3
+from botocore.exceptions import ClientError
 import os
 from datalake.common.errors import InsufficientConfiguration
 import logging
@@ -42,43 +41,55 @@ def from_config(cls):
 
     def _prepare_connection(self, connection):
         self.logger.info("Preparing connection...")
-        region = os.environ.get('AWS_REGION')
+        # boto2 defaulted to us-east-1, maintaining consistency across all AWS services
+        region = os.environ.get('AWS_REGION', 'us-east-1')
         if connection:
+            # When connection is provided from outside, we need to ensure _client is set
             self._connection = connection
-        elif region:
-            self._connection = boto.dynamodb2.connect_to_region(region)
+
+            # Check if _connection has a client attribute (added in our tests)
+            # or create a new client if it doesn't
+            if hasattr(connection, 'client'):
+                self._client = connection.client
+            else:
+                self._client = boto3.client('dynamodb', region_name=region)
         else:
-            msg = 'Please provide a connection or configure a region'
-            raise InsufficientConfiguration(msg)
+            self._connection = boto3.resource('dynamodb', region_name=region)
+            self._client = boto3.client('dynamodb', region_name=region)
 
     @memoized_property
     def _table(self):
-        return Table(self.table_name, connection=self._connection)
-    
+        return self._connection.Table(self.table_name)
+
     @memoized_property
     def _latest_table(self):
-        return Table(self.latest_table_name, connection=self._connection)
+        return self._connection.Table(self.latest_table_name)
 
     def store(self, record):
         try:
-            self._table.put_item(data=record)
-        except ConditionalCheckFailedException:
-            # Tolerate duplicate stores
-            pass
+            self._table.put_item(Item=record)
+        except ClientError as e:
+            if e.response['Error']['Code'] == 'ConditionalCheckFailedException':
+                # Tolerate duplicate stores
+                pass
+            else:
+                raise
         if self.latest_table_name:
             self.store_latest(record)
 
     def update(self, record):
-        self._table.put_item(data=record, overwrite=True)
+        self._table.put_item(Item=record)
 
     def store_latest(self, record):
         """
         Record must utilize AttributeValue syntax
               for the conditional put.
         """
-        condition_expression = " attribute_not_exists(what_where_key) OR metadata.#metadata_start <= :new_start"
+        # Boto3 requires different parameter naming: condition_expression -> ConditionExpression
+        condition_expression = "attribute_not_exists(what_where_key) OR metadata.#metadata_start <= :new_start"
+
         expression_attribute_values = {
-            ':new_start': {'N': str(record['metadata']['start'])}
+            ':new_start': {'N': str(record['metadata']['start'])}  # Must use typed dict here
         }
 
         # aliases for DynamoDB reserved names.
@@ -132,17 +143,21 @@ def store_latest(self, record):
         }
         self.logger.info(f"Attempting to store record: {record}")
         try:
-            self._connection.put_item(
-                table_name=self.latest_table_name,
-                item=record,
-                condition_expression=condition_expression,
-                expression_attribute_names=expression_attribute_names,
-                expression_attribute_values=expression_attribute_values,
+            self._client.put_item(
+                TableName=self.latest_table_name,
+                Item=record,
+                ConditionExpression=condition_expression,
+                ExpressionAttributeNames=expression_attribute_names,
+                ExpressionAttributeValues=expression_attribute_values,
             )
             self.logger.info("Record stored successfully.")
-        except ConditionalCheckFailedException:
-            self.logger.debug(f"Condition not met for record {record},"
-                              "no operation was performed.")
-        except Exception as e:
-            self.logger.error(f"Error occurred while attempting {record}: {str(e)}")
+        except ClientError as e:
+            # Handle conditional check failures (expected when record is older than existing one)
+            if e.response['Error']['Code'] == 'ConditionalCheckFailedException':
+                self.logger.debug(f"Condition not met for record {record}, "
+                                  "no operation was performed.")
+            else:
+                # All other AWS/boto3 errors - log and re-raise
+                self.logger.error(f"Error occurred while attempting {record}: {str(e)}")
+                raise