repo
stringlengths
7
90
file_url
stringlengths
81
315
file_path
stringlengths
4
228
content
stringlengths
0
32.8k
language
stringclasses
1 value
license
stringclasses
7 values
commit_sha
stringlengths
40
40
retrieved_at
stringdate
2026-01-04 14:38:15
2026-01-05 02:33:18
truncated
bool
2 classes
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline-sfn/tests/test_stack.py
modules/analysis/rosbag-image-pipeline-sfn/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" # Unload the app import so that subsequent tests don't reuse if "stack" in sys.modules: del sys.modules["stack"] def test_stack(stack_defaults): import stack app = cdk.App() project_name = "test-project" dep_name = "test-deployment" mod_name = "test-module" rosbag_stack = stack.TemplateStack( scope=app, id=f"{project_name}-{dep_name}-{mod_name}", project_name=project_name, deployment_name=dep_name, module_name=mod_name, hash="foobar", vpc_id="vpc-id", source_bucket_name="raw-bucket", target_bucket_name="intermediate-bucket", logs_bucket_name="logs-bucket", artifacts_bucket_name="artifacts-bucket", private_subnet_ids='["foo","bar"]', emr_job_exec_role_arn="dummy-emr-role", emr_app_id="012345678", detection_ddb_name="scene-metadata-table", on_demand_job_queue_arn="arn:aws:batch:us-west-2:123456789101:job-queue/addf-example", fargate_job_queue_arn="arn:aws:batch:us-west-2:123456789101:job-queue/addf-example", parquet_batch_job_def_arn="dummyarn", png_batch_job_def_arn="dummyarn", object_detection_image_uri="dummyuri", object_detection_role_arn="dummyrole", object_detection_job_concurrency=10, object_detection_instance_type="ml.2xlarge", lane_detection_image_uri="dummyuri", lane_detection_role_arn="dummyrole", lane_detection_job_concurrency=10, lane_detection_instance_type="ml.2xlarge", file_suffix=".bag", desired_encoding="bgr8", yolo_model="yolov5s", stack_description="Testing", image_topics='["foo", "bar"]', sensor_topics='["foo", "bar"]', env=cdk.Environment( account=(os.environ["CDK_DEFAULT_ACCOUNT"]), region=(os.environ["CDK_DEFAULT_REGION"]), ), ) template = Template.from_stack(rosbag_stack) template.resource_count_is("AWS::DynamoDB::Table", 1) template.resource_count_is("AWS::IAM::Role", 3) template.resource_count_is("AWS::StepFunctions::StateMachine", 1)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline-sfn/tests/__init__.py
modules/analysis/rosbag-image-pipeline-sfn/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline-sfn/lambda/create-batch-of-drives/src/lambda_function.py
modules/analysis/rosbag-image-pipeline-sfn/lambda/create-batch-of-drives/src/lambda_function.py
import logging import os import typing import boto3 from boto3.dynamodb.conditions import Key if typing.TYPE_CHECKING: from mypy_boto3_dynamodb.service_resource import DynamoDBServiceResource from mypy_boto3_s3.client import S3Client logger = logging.getLogger() logger.setLevel("DEBUG") dynamodb_resource = boto3.resource("dynamodb") s3_client = boto3.client("s3") DYNAMODB_TABLE = os.environ["DYNAMODB_TABLE"] FILE_SUFFIX = os.environ["FILE_SUFFIX"] def add_drives_to_batch( table: str, batch_id: str, drives_to_process: typing.Dict[str, dict], file_suffix: str, s3_client: "S3Client", ) -> int: """ Lists files with file_suffix for each prefix in drives_to_process and adds each file to DynamoDB for tracking @param table: dynamo tracking table @param batch_id: state machine execution id @param drives_to_process: { "drive1": {"bucket": "addf-example-dev-raw-bucket-xyz", "prefix": "rosbag-scene-detection/drive1/"}, "drive2": {"bucket": "addf-example-dev-raw-bucket-xyz", "prefix": "rosbag-scene-detection/drive2/"}, } @param file_suffix: ".bag" @param s3_client: type boto3.client('s3') """ drives_and_files = {} files_in_batch = 0 for drive_id, s3_path in drives_to_process.items(): files = get_drive_files( src_bucket=s3_path["bucket"], src_prefix=s3_path["prefix"], file_suffix=file_suffix, s3_client=s3_client, ) drives_and_files[drive_id] = {"files": files, "bucket": s3_path["bucket"]} files_in_batch += len(files) logger.info(f"files_in_batch {files_in_batch}") batch_write_files_to_dynamo(table, drives_and_files, batch_id) return files_in_batch def get_drive_files(src_bucket: str, src_prefix: str, file_suffix: str, s3_client: "S3Client") -> typing.List[str]: """For a given bucket, prefix, and suffix, lists all files found on S3 and returns a list of the files @param drive_id: @param src_bucket: @param src_prefix: @param file_suffix: @param s3_client: @return: """ MAX_KEYS = 1000 logger.info(src_bucket) logger.info(src_prefix) file_response = s3_client.list_objects_v2(Bucket=src_bucket, Prefix=src_prefix, MaxKeys=MAX_KEYS, Delimiter="/") logger.info(file_response) file_next_continuation = file_response.get("NextContinuationToken") files = [x["Key"] for x in file_response.get("Contents", []) if x["Key"].endswith(file_suffix)] while file_next_continuation is not None: file_response = s3_client.list_objects_v2( Bucket=src_bucket, Prefix=src_prefix, MaxKeys=MAX_KEYS, Delimiter="/", ContinuationToken=file_next_continuation, ) file_next_continuation = file_response.get("NextContinuationToken") files += [x["Key"] for x in file_response.get("Contents", [])] logger.info(files) return files def batch_write_files_to_dynamo( table: "DynamoDBServiceResource", drives_and_files: typing.Dict[str, typing.List[str]], batch_id: str, ) -> None: with table.batch_writer() as writer: idx = 0 for drive_id, files in drives_and_files.items(): for file in files["files"]: item = { "drive_id": drive_id, "file_id": file.split("/")[-1], "s3_bucket": files["bucket"], "s3_key": file, "pk": batch_id, "sk": str(idx), } logger.info(item) writer.put_item(Item=item) idx += 1 def lambda_handler(event: typing.Dict[str, typing.Any], context: typing.Any) -> int: drives_to_process = event["DrivesToProcess"] execution_id = event["ExecutionID"] table = dynamodb_resource.Table(DYNAMODB_TABLE) files_in_batch = table.query( KeyConditionExpression=Key("pk").eq(execution_id), Select="COUNT", )["Count"] if files_in_batch > 0: logger.info("Batch Id already exists in tracking table - using existing batch") return {"BatchSize": files_in_batch, "ExecutionID": execution_id} logger.info("New Batch Id - collecting unprocessed drives from S3 and adding to the batch") files_in_batch = add_drives_to_batch( table=table, drives_to_process=drives_to_process, batch_id=execution_id, file_suffix=FILE_SUFFIX, s3_client=s3_client, ) if files_in_batch > 10_000: raise RuntimeError("Batch Size cannot exceed 10,000") return {"BatchSize": files_in_batch, "ExecutionID": execution_id}
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline-sfn/emr-scripts/detect_scenes.py
modules/analysis/rosbag-image-pipeline-sfn/emr-scripts/detect_scenes.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import argparse import json import sys from typing import Any, Dict, List import boto3 import pyspark.sql.functions as func from pyspark.sql import SparkSession, Window from pyspark.sql.functions import ( aggregate, col, concat, count, first, from_json, lit, split, ) from pyspark.sql.types import ( ArrayType, DoubleType, IntegerType, StringType, StructField, StructType, ) obj_schema = StructType( [ StructField("_c0", IntegerType(), True), StructField("xmin", DoubleType(), True), StructField("ymin", DoubleType(), True), StructField("xmax", DoubleType(), True), StructField("ymax", DoubleType(), True), StructField("confidence", DoubleType(), True), StructField("class", IntegerType(), True), StructField("name", StringType(), True), StructField("source_image", StringType(), True), ] ) def parse_arguments(args): parser = argparse.ArgumentParser() parser.add_argument("--batch-metadata-table-name", required=True) parser.add_argument("--batch-id", required=True) parser.add_argument("--input-bucket", required=False) parser.add_argument("--output-bucket", required=True) parser.add_argument("--output-dynamo-table", required=True) parser.add_argument("--region", required=True) parser.add_argument("--image-topics", required=False) return parser.parse_args(args=args) def form_object_in_lane_df(obj_df, lane_df): obj_lane_df = obj_df.join(lane_df, on="source_image", how="inner") obj_lane_df = ( obj_lane_df.withColumn( "pixel_rows_at_bottom_corner", obj_lane_df["lanes"].getItem(obj_lane_df.ymax.cast(IntegerType())), ) .drop("lanes") .drop("ymax") ) obj_lane_df = obj_lane_df.withColumn( "sum_of_pixels_intensities_at_bottom_corner_rows", aggregate("pixel_rows_at_bottom_corner", lit(0), lambda acc, x: acc + x), ).drop("pixel_rows_at_bottom_corner") obj_lane_df = obj_lane_df.filter(obj_lane_df.sum_of_pixels_intensities_at_bottom_corner_rows != 0) obj_lane_df = obj_lane_df.withColumn("source_image_split", split("source_image", "_")) obj_lane_df = obj_lane_df.withColumn( "Time", concat( obj_lane_df["source_image_split"].getItem(2), lit("."), obj_lane_df["source_image_split"].getItem(3).substr(1, 2), ).cast(DoubleType()), ) return obj_lane_df def get_batch_file_metadata(table_name, batch_id, region): dynamodb = boto3.resource("dynamodb", region_name=region) table = dynamodb.Table(table_name) response = table.query(KeyConditions={"pk": {"AttributeValueList": [batch_id], "ComparisonOperator": "EQ"}}) data = response["Items"] while "LastEvaluatedKey" in response: response = table.query(ExclusiveStartKey=response["LastEvaluatedKey"]) data.update(response["Items"]) return data def load_obj_detection(spark, batch_metadata, image_topics): path_list = [] def _process( image_topics: List[str], path_list: List[str], item: Dict[str, Any], resizied_image_dir: str, ) -> None: for t in image_topics: path_list.append( f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_obj_dets/all_predictions.csv" ) if t in resizied_image_dir else None for item in batch_metadata: for resizied_image_dir in item["resized_image_dirs"]: if image_topics and len(image_topics) > 0: _process(image_topics, path_list, item, resizied_image_dir) else: path_list.append( f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_obj_dets/all_predictions.csv" ) def remove_space(name): return name.replace(" ", "_") remove_space_udf = func.udf(remove_space, StringType()) df = spark.read.schema(obj_schema).option("header", True).csv(path_list) df2 = df.withColumn("name", remove_space_udf(df.name)) print("Number of rows in Object Detection dataframe") print(df2.count()) return df2 def load_lane_detection(spark, batch_metadata): first_item = batch_metadata[0] first_path_prefix = first_item["resized_image_dirs"][0] first_path = f"s3://{first_item['raw_image_bucket']}/{first_path_prefix}_post_lane_dets/lanes.csv" first_item_split = first_item["s3_key"].rpartition("/") bag_file_prefix = first_item_split[0] bag_file = first_item_split[2].split(".")[0] df = ( spark.read.option("header", True) .csv(first_path) .withColumn("bag_file_bucket", lit(first_item["s3_bucket"])) .withColumn("bag_file", lit(bag_file)) .withColumn("bag_file_prefix", lit(bag_file_prefix)) ) for i in range(1, len(batch_metadata)): item = batch_metadata[i] for resizied_image_dir in item["resized_image_dirs"]: path = f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_lane_dets/lanes.csv" item_split = item["s3_key"].rpartition("/") bag_file_prefix = item_split[0] bag_file = item_split[2].split(".")[0] df.union( spark.read.option("header", True) .csv(path) .withColumn("bag_file_bucket", lit(item["s3_bucket"])) .withColumn("bag_file", lit(bag_file)) .withColumn("bag_file_prefix", lit(bag_file_prefix)) ) lane_schema = ArrayType(ArrayType(IntegerType()), False) df = df.withColumn("lanes", from_json(col("lanes"), lane_schema)) return df def write_results_s3(dfs: Dict[str, Any], table_name: str, output_bucket: str, partition_cols=[]): for key, df in dfs.items(): s3_path = f"s3://{output_bucket}/{table_name}/{key}" df.write.mode("append").partitionBy(*partition_cols).parquet(s3_path) def write_results_dynamo(dfs, output_dynamo_table, region): for key, df in dfs.items(): print(f"Writing {key} to DDB") df.write.mode("append").option("tableName", output_dynamo_table).option("region", region).format( "dynamodb" ).save() def summarize_obj_in_lane_scenes(obj_lane_df, image_topics, obj_type): obj_lane_df = obj_lane_df.filter(obj_lane_df.name == obj_type) obj_lane_df = obj_lane_df.groupby("source_image").agg( count("name").alias(f"num_{obj_type}_in_lane"), first("Time").alias("Time"), first("bag_file_bucket").alias("bag_file_bucket"), first("bag_file").alias("bag_file"), first("bag_file_prefix").alias("bag_file_prefix"), ) win = Window.orderBy("Time").partitionBy("bag_file_bucket", "bag_file", "bag_file_prefix") obj_lane_df = ( obj_lane_df.withColumn( f"num_{obj_type}_in_lane_lag1", func.lag(func.col(f"num_{obj_type}_in_lane"), 1, 0).over(win), ) .withColumn( f"num_{obj_type}_in_lane_lead1", func.lead(func.col(f"num_{obj_type}_in_lane"), 1, 0).over(win), ) .filter(f"num_{obj_type}_in_lane_lag1 == 0 or num_{obj_type}_in_lane_lead1 ==0") ) scene_state_udf = func.udf( lambda num, lag: "start" if num > 0 and lag is None else ("end" if num == 0 and lag > 0 else None), StringType(), ) obj_in_lane_scenes_df = ( obj_lane_df.withColumn( "scene_state", scene_state_udf( obj_lane_df[f"num_{obj_type}_in_lane"], obj_lane_df[f"num_{obj_type}_in_lane_lag1"], ), ) .withColumn("end_time", func.lead(func.col("Time"), 1).over(win)) .filter(f"num_{obj_type}_in_lane_lag1 ==0") .withColumnRenamed("Time", "start_time") .withColumnRenamed(f"num_{obj_type}_in_lane", f"num_{obj_type}_in_lane_start") .select( "bag_file", "bag_file_bucket", "bag_file_prefix", "start_time", "end_time", f"num_{obj_type}_in_lane_start", ) .withColumn( "scene_id", func.concat( func.col("bag_file"), func.lit(f"_{obj_type}InLane_"), func.col("start_time"), ), ) .withColumn("scene_length", func.col("end_time") - func.col("start_time")) .withColumn("topics_analyzed", func.lit(",".join(image_topics))) ) return obj_in_lane_scenes_df def main( batch_metadata_table_name, batch_id, output_bucket, output_dynamo_table, spark, region, image_topics, ): # Load files to process batch_metadata = get_batch_file_metadata(table_name=batch_metadata_table_name, batch_id=batch_id, region=region) if image_topics: image_topics = json.loads(image_topics) image_topics = [topic.replace("/", "_") for topic in image_topics if image_topics] # Load topic data from s3 and union obj_df = load_obj_detection(spark, batch_metadata=batch_metadata, image_topics=image_topics) lane_df = load_lane_detection( spark, batch_metadata=batch_metadata, ) obj_lane_df = form_object_in_lane_df(obj_df, lane_df) dfs = {} dfs["car"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "car") dfs["truck"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "truck") dfs["traffic_light"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "traffic_light") dfs["train"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "train") dfs["bus"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "bus") dfs["motorcycle"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "motorcycle") dfs["stop_sign"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "stop_sign") dfs["fire_hydrant"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "fire_hydrant") write_results_s3( dfs, table_name="scene_detections", output_bucket=output_bucket, partition_cols=["bag_file"], ) write_results_dynamo(dfs, output_dynamo_table, region) if __name__ == "__main__": spark = SparkSession.builder.appName("scene-detection").getOrCreate() sc = spark.sparkContext arguments = parse_arguments(sys.argv[1:]) batch_metadata_table_name = arguments.batch_metadata_table_name batch_id = arguments.batch_id output_bucket = arguments.output_bucket output_dynamo_table = arguments.output_dynamo_table region = arguments.region image_topics = arguments.image_topics if arguments.image_topics else None main( batch_metadata_table_name, batch_id, output_bucket, output_dynamo_table, spark, region, image_topics, ) sc.stop()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/stack.py
modules/analysis/rosbag-image-pipeline/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, List, cast import aws_cdk.aws_ec2 as ec2 import aws_cdk.aws_iam as iam import cdk_nag from aws_cdk import Aspects, Duration, RemovalPolicy, Stack, Tags from aws_cdk import aws_dynamodb as dynamo from cdk_nag import NagPackSuppression, NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class AwsBatchPipeline(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, vpc_id: str, mwaa_exec_role: str, bucket_access_policy: str, object_detection_role: str, lane_detection_role: str, job_queues: List[str], job_definitions: List[str], stack_description: str, **kwargs: Any, ) -> None: super().__init__( scope, id, description=stack_description, **kwargs, ) self.project_name = project_name self.deployment_name = deployment_name self.module_name = module_name self.vpc_id = vpc_id self.mwaa_exec_role = mwaa_exec_role self.bucket_access_policy = bucket_access_policy Tags.of(scope=cast(IConstruct, self)).add( key="Deployment", value="aws", ) dep_mod = f"{self.project_name}-{self.deployment_name}-{self.module_name}" # DYNAMODB TRACKING TABLE self.tracking_table_name = f"{dep_mod}-drive-tracking" tracking_partition_key = "pk" # batch_id or drive_id tracking_sort_key = "sk" # batch_id / array_index_id or drive_id / file_part dynamo.Table( self, self.tracking_table_name, table_name=self.tracking_table_name, partition_key=dynamo.Attribute(name=tracking_partition_key, type=dynamo.AttributeType.STRING), sort_key=dynamo.Attribute(name=tracking_sort_key, type=dynamo.AttributeType.STRING), billing_mode=dynamo.BillingMode.PAY_PER_REQUEST, removal_policy=RemovalPolicy.DESTROY, point_in_time_recovery=True, stream=dynamo.StreamViewType.NEW_AND_OLD_IMAGES, ) # Create Dag IAM Role and policy policy_statements = [ iam.PolicyStatement( actions=["dynamodb:*"], effect=iam.Effect.ALLOW, resources=[f"arn:{self.partition}:dynamodb:{self.region}:{self.account}:table/{dep_mod}*"], ), iam.PolicyStatement( actions=["ecr:*"], effect=iam.Effect.ALLOW, resources=[f"arn:{self.partition}:ecr:{self.region}:{self.account}:repository/{dep_mod}*"], ), iam.PolicyStatement( actions=[ "batch:UntagResource", "batch:DeregisterJobDefinition", "batch:TerminateJob", "batch:CancelJob", "batch:SubmitJob", "batch:RegisterJobDefinition", "batch:TagResource", ], effect=iam.Effect.ALLOW, resources=[ *job_queues, *job_definitions, f"arn:{self.partition}:batch:{self.region}:{self.account}:job/*", ], ), iam.PolicyStatement( actions=[ "iam:PassRole", ], effect=iam.Effect.ALLOW, resources=[x for x in [object_detection_role, lane_detection_role] if x is not None], ), iam.PolicyStatement( actions=[ "batch:Describe*", "batch:List*", ], effect=iam.Effect.ALLOW, resources=[ "*", ], ), iam.PolicyStatement( actions=["s3:GetObject", "s3:GetObjectAcl", "s3:ListBucket"], effect=iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:s3:::{project_name}-*", f"arn:{self.partition}:s3:::{project_name}-*/*", ], ), ] dag_document = iam.PolicyDocument(statements=policy_statements) dag_role_name = f"{dep_mod}-dag-{self.region}" self.dag_role = iam.Role( self, f"dag-role-{dep_mod}", assumed_by=iam.CompositePrincipal( iam.ArnPrincipal(self.mwaa_exec_role), ), inline_policies={"DagPolicyDocument": dag_document}, managed_policies=[ iam.ManagedPolicy.from_managed_policy_arn( self, id="fullaccess", managed_policy_arn=self.bucket_access_policy ), iam.ManagedPolicy.from_aws_managed_policy_name("AmazonSageMakerFullAccess"), ], role_name=dag_role_name, max_session_duration=Duration.hours(12), path="/", ) # Sagemaker Security Group self.vpc = ec2.Vpc.from_lookup( self, "VPC", vpc_id=self.vpc_id, ) self.sm_sg = ec2.SecurityGroup( self, "SagemakerJobsSG", vpc=self.vpc, allow_all_outbound=True, description="Sagemaker Processing Jobs SG", ) self.sm_sg.add_ingress_rule(peer=self.sm_sg, connection=ec2.Port.all_traffic()) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) NagSuppressions.add_stack_suppressions( self, apply_to_nested_stacks=True, suppressions=[ NagPackSuppression( **{ "id": "AwsSolutions-IAM4", "reason": "Managed Policies are for service account roles only", } ), NagPackSuppression( **{ "id": "AwsSolutions-IAM5", "reason": "Resource access restriced to ADDF resources", } ), ], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/app.py
modules/analysis/rosbag-image-pipeline/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import json import os from aws_cdk import App, CfnOutput, Environment from stack import AwsBatchPipeline project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" dag_id = os.getenv(_param("DAG_ID")) vpc_id = os.getenv(_param("VPC_ID")) private_subnet_ids = json.loads(os.getenv(_param("PRIVATE_SUBNET_IDS"))) # type: ignore mwaa_exec_role = os.getenv(_param("MWAA_EXEC_ROLE")) full_access_policy = os.getenv(_param("FULL_ACCESS_POLICY_ARN")) source_bucket_name = os.getenv(_param("SOURCE_BUCKET")) target_bucket_name = os.getenv(_param("INTERMEDIATE_BUCKET")) dag_bucket_name = os.getenv(_param("DAG_BUCKET_NAME")) log_bucket_name = os.getenv(_param("LOGS_BUCKET_NAME")) detection_ddb_name = os.getenv(_param("ROSBAG_SCENE_METADATA_TABLE")) on_demand_job_queue = os.getenv(_param("ON_DEMAND_JOB_QUEUE_ARN")) spot_job_queue = os.getenv(_param("SPOT_JOB_QUEUE_ARN")) fargate_job_queue = os.getenv(_param("FARGATE_JOB_QUEUE_ARN")) parquet_batch_job_def_arn = os.getenv(_param("PARQUET_BATCH_JOB_DEF_ARN")) png_batch_job_def_arn = os.getenv(_param("PNG_BATCH_JOB_DEF_ARN")) object_detection_image_uri = os.getenv(_param("OBJECT_DETECTION_IMAGE_URI")) object_detection_role = os.getenv(_param("OBJECT_DETECTION_IAM_ROLE")) object_detection_job_concurrency = int(os.getenv(_param("OBJECT_DETECTION_JOB_CONCURRENCY"), 10)) object_detection_instance_type = os.getenv(_param("OBJECT_DETECTION_INSTANCE_TYPE"), "ml.m5.xlarge") lane_detection_image_uri = os.getenv(_param("LANE_DETECTION_IMAGE_URI")) lane_detection_role = os.getenv(_param("LANE_DETECTION_IAM_ROLE")) lane_detection_job_concurrency = int(os.getenv(_param("LANE_DETECTION_JOB_CONCURRENCY"), 5)) lane_detection_instance_type = os.getenv(_param("LANE_DETECTION_INSTANCE_TYPE"), "ml.p3.2xlarge") file_suffix = os.getenv(_param("FILE_SUFFIX"), ".bag") desired_encoding = os.getenv(_param("DESIRED_ENCODING"), "bgr8") yolo_model = os.getenv(_param("YOLO_MODEL"), "yolo11s") image_topics = os.getenv(_param("IMAGE_TOPICS")) sensor_topics = os.getenv(_param("SENSOR_TOPICS")) if not png_batch_job_def_arn: raise ValueError("missing input parameter png-batch-job-def-arn") if not parquet_batch_job_def_arn: raise ValueError("missing input parameter parquet-batch-job-def-arn") if not object_detection_role: raise ValueError("missing input parameter object-detection-iam-role") if not object_detection_image_uri: raise ValueError("missing input parameter object-detection-image-uri") if not lane_detection_role: raise ValueError("missing input parameter lane-detection-iam-role") if not lane_detection_image_uri: raise ValueError("missing input parameter lane-detection-image-uri") if not vpc_id: raise ValueError("missing input parameter vpc-id") if not private_subnet_ids: raise ValueError("missing input parameter private-subnet-ids") if not mwaa_exec_role: raise ValueError("MWAA Execution Role is missing.") if not full_access_policy: raise ValueError("S3 Full Access Policy ARN is missing.") if not on_demand_job_queue and not spot_job_queue and not fargate_job_queue: raise ValueError("Requires at least one job queue.") def generate_description() -> str: soln_id = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_ID", None) soln_name = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_NAME", None) soln_version = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_VERSION", None) desc = "(SO9154) Autonomous Driving Data Framework (ADDF) - rosbag-image-pipeline" if soln_id and soln_name and soln_version: desc = f"({soln_id}) {soln_name}. Version {soln_version}" elif soln_id and soln_name: desc = f"({soln_id}) {soln_name}" return desc app = App() stack = AwsBatchPipeline( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, vpc_id=vpc_id, mwaa_exec_role=mwaa_exec_role, bucket_access_policy=full_access_policy, object_detection_role=object_detection_role, lane_detection_role=lane_detection_role, job_queues=[x for x in [fargate_job_queue, spot_job_queue, on_demand_job_queue] if x is not None], job_definitions=[x for x in [png_batch_job_def_arn, parquet_batch_job_def_arn] if x is not None], stack_description=generate_description(), env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "PrivateSubnetIds": private_subnet_ids, "DagId": dag_id, "SecurityGroupId": stack.sm_sg.security_group_id, "DagRoleArn": stack.dag_role.role_arn, "DynamoDbTableName": stack.tracking_table_name, "DetectionsDynamoDBName": detection_ddb_name, "SourceBucketName": source_bucket_name, "TargetBucketName": target_bucket_name, "DagBucketName": dag_bucket_name, "LogsBucketName": log_bucket_name, "OnDemandJobQueueArn": on_demand_job_queue, "SpotJobQueueArn": spot_job_queue, "FargateJobQueueArn": fargate_job_queue, "ParquetBatchJobDefArn": parquet_batch_job_def_arn, "PngBatchJobDefArn": png_batch_job_def_arn, "ObjectDetectionImageUri": object_detection_image_uri, "ObjectDetectionRole": object_detection_role, "ObjectDetectionJobConcurrency": object_detection_job_concurrency, "ObjectDetectionInstanceType": object_detection_instance_type, "LaneDetectionImageUri": lane_detection_image_uri, "LaneDetectionRole": lane_detection_role, "LaneDetectionJobConcurrency": lane_detection_job_concurrency, "LaneDetectionInstanceType": lane_detection_instance_type, "FileSuffix": file_suffix, "DesiredEncoding": desired_encoding, "YoloModel": yolo_model, "ImageTopics": json.loads(image_topics), # type: ignore "SensorTopics": json.loads(sensor_topics), # type: ignore } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/test_app.py
modules/analysis/rosbag-image-pipeline/tests/test_app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys from json import JSONDecodeError import pytest @pytest.fixture(scope="function") def stack_defaults(): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project" os.environ["SEEDFARMER_DEPLOYMENT_NAME"] = "test-deployment" os.environ["SEEDFARMER_MODULE_NAME"] = "test-module" os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" os.environ["SEEDFARMER_PARAMETER_DAG_ID"] = "dag-id" os.environ["SEEDFARMER_PARAMETER_VPC_ID"] = "vpc-id" os.environ["SEEDFARMER_PARAMETER_PRIVATE_SUBNET_IDS"] = '["subnet-12345", "subnet-54321"]' os.environ["SEEDFARMER_PARAMETER_MWAA_EXEC_ROLE"] = "mwaa-exec-role" os.environ["SEEDFARMER_PARAMETER_FULL_ACCESS_POLICY_ARN"] = "full-access-policy-arn" os.environ["SEEDFARMER_PARAMETER_SOURCE_BUCKET"] = "source-bucket" os.environ["SEEDFARMER_PARAMETER_INTERMEDIATE_BUCKET"] = "intermediate-bucket" os.environ["SEEDFARMER_PARAMETER_ON_DEMAND_JOB_QUEUE_ARN"] = "on-demand-job-queue-arn" os.environ["SEEDFARMER_PARAMETER_SPOT_JOB_QUEUE_ARN"] = "spot-job-queue-arn" os.environ["SEEDFARMER_PARAMETER_FARGATE_JOB_QUEUE_ARN"] = "fargate-job-queue-arn" os.environ["SEEDFARMER_PARAMETER_PARQUET_BATCH_JOB_DEF_ARN"] = "parquet-batch-job-def-arn" os.environ["SEEDFARMER_PARAMETER_PNG_BATCH_JOB_DEF_ARN"] = "png-batch-job-def-arn" os.environ["SEEDFARMER_PARAMETER_OBJECT_DETECTION_IMAGE_URI"] = "object-detection-image-uri" os.environ["SEEDFARMER_PARAMETER_OBJECT_DETECTION_IAM_ROLE"] = "object-detection-iam-role" os.environ["SEEDFARMER_PARAMETER_LANE_DETECTION_IMAGE_URI"] = "lane-detection-image-uri" os.environ["SEEDFARMER_PARAMETER_LANE_DETECTION_IAM_ROLE"] = "lane-detection-iam-role" os.environ["SEEDFARMER_PARAMETER_IMAGE_TOPICS"] = "{}" os.environ["SEEDFARMER_PARAMETER_SENSOR_TOPICS"] = "{}" # Unload the app import so that subsequent tests don't reuse if "app" in sys.modules: del sys.modules["app"] def test_app(stack_defaults): import app # noqa: F401 def test_png_batch_job_def_arn(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_PNG_BATCH_JOB_DEF_ARN"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter png-batch-job-def-arn" in str(e) def test_parquet_batch_job_def_arn(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_PARQUET_BATCH_JOB_DEF_ARN"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter parquet-batch-job-def-arn" in str(e) def test_object_detection_role(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_OBJECT_DETECTION_IAM_ROLE"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter object-detection-iam-role" in str(e) def test_object_detection_image_uri(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_LANE_DETECTION_IMAGE_URI"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter lane-detection-image-uri" in str(e) def test_lane_detection_role(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_LANE_DETECTION_IAM_ROLE"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter lane-detection-iam-role" in str(e) def test_lane_detection_image_uri(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_LANE_DETECTION_IMAGE_URI"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter lane-detection-image-uri" in str(e) def test_vpc_id(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_VPC_ID"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter vpc-id" in str(e) def test_private_subnet_ids(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_PRIVATE_SUBNET_IDS"] with pytest.raises(Exception) as e: import app # noqa: F401 assert "missing input parameter private-subnet-ids" in str(e) def test_mwaa_exec_role(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_MWAA_EXEC_ROLE"] with pytest.raises(ValueError) as e: import app # noqa: F401 assert "MWAA Execution Role is missing." in str(e) def test_full_access_policy(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_FULL_ACCESS_POLICY_ARN"] with pytest.raises(ValueError) as e: import app # noqa: F401 assert "S3 Full Access Policy ARN is missing." in str(e) def test_no_queue_provided(): del os.environ["SEEDFARMER_PARAMETER_ON_DEMAND_JOB_QUEUE_ARN"] del os.environ["SEEDFARMER_PARAMETER_SPOT_JOB_QUEUE_ARN"] del os.environ["SEEDFARMER_PARAMETER_FARGATE_JOB_QUEUE_ARN"] with pytest.raises(ValueError) as e: import app # noqa: F401 assert "Requires at least one job queue." in str(e) def test_image_topics_no_json(stack_defaults): os.environ["SEEDFARMER_PARAMETER_IMAGE_TOPICS"] = "no json" with pytest.raises(JSONDecodeError): import app # noqa: F401 def test_sensor_topics_no_json(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SENSOR_TOPICS"] = "no json" with pytest.raises(JSONDecodeError): import app # noqa: F401 def test_solution_description(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] = "v1.0.0" import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST. Version v1.0.0" def test_solution_description_no_version(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" del os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST"
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/test_stack.py
modules/analysis/rosbag-image-pipeline/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" # Unload the app import so that subsequent tests don't reuse if "stack" in sys.modules: del sys.modules["stack"] def iam_policy() -> dict: return { "Policies": [ { "PolicyDocument": { "Statement": [ { "Action": "dynamodb:*", "Effect": "Allow", "Resource": {'Fn::Join': ['', ['arn:', {'Ref': 'AWS::Partition'}, ':dynamodb:us-east-1:111111111111:table/test-project-test-deployment-test-module*']]} }, { "Action": "ecr:*", "Effect": "Allow", "Resource": {'Fn::Join': ['', ['arn:', {'Ref': 'AWS::Partition'}, ':ecr:us-east-1:111111111111:repository/test-project-test-deployment-test-module*']]} }, { "Action": [ "batch:UntagResource", "batch:DeregisterJobDefinition", "batch:TerminateJob", "batch:CancelJob", "batch:SubmitJob", "batch:RegisterJobDefinition", "batch:TagResource", ], "Effect": "Allow", "Resource": [ "job-queue-1", "job-queue-2", "job-def-1", "job-def-2", {'Fn::Join': ['', ['arn:', {'Ref': 'AWS::Partition'}, ':batch:us-east-1:111111111111:job/*']]} ], }, { "Action": "iam:PassRole", "Effect": "Allow", "Resource": [ "obj-detection-role", "lane-detection-role", ], }, { "Action": ["batch:Describe*", "batch:List*"], "Effect": "Allow", "Resource": "*", }, { "Action": [ "s3:GetObject", "s3:GetObjectAcl", "s3:ListBucket", ], "Effect": "Allow", "Resource": [ {'Fn::Join': ['', ['arn:', {'Ref': 'AWS::Partition'}, ':s3:::test-project-*']]}, {'Fn::Join': ['', ['arn:', {'Ref': 'AWS::Partition'}, ':s3:::test-project-*/*']]} ] }, ] } } ], } def test_stack(stack_defaults): import stack app = cdk.App() project_name = "test-project" dep_name = "test-deployment" mod_name = "test-module" rosbag_stack = stack.AwsBatchPipeline( scope=app, id=f"{project_name}-{dep_name}-{mod_name}", project_name=project_name, deployment_name=dep_name, module_name=mod_name, vpc_id="vpc-id", mwaa_exec_role="mwaa-exec-role", bucket_access_policy="bucket-access-policy", object_detection_role="obj-detection-role", lane_detection_role="lane-detection-role", job_queues=["job-queue-1", "job-queue-2"], job_definitions=["job-def-1", "job-def-2"], stack_description="Testing", env=cdk.Environment( account=(os.environ["CDK_DEFAULT_ACCOUNT"]), region=(os.environ["CDK_DEFAULT_REGION"]), ), ) template = Template.from_stack(rosbag_stack) template.resource_count_is("AWS::DynamoDB::Table", 1) template.resource_count_is("AWS::IAM::Role", 1) template.has_resource_properties( "AWS::IAM::Role", { "AssumeRolePolicyDocument": { "Statement": [ { "Action": "sts:AssumeRole", "Effect": "Allow", "Principal": {"AWS": "mwaa-exec-role"}, } ], "Version": "2012-10-17", }, "ManagedPolicyArns": [ "bucket-access-policy", { "Fn::Join": [ "", [ "arn:", {"Ref": "AWS::Partition"}, ":iam::aws:policy/AmazonSageMakerFullAccess", ], ] }, ], }, ) template.has_resource_properties("AWS::IAM::Role", iam_policy())
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/conftest.py
modules/analysis/rosbag-image-pipeline/tests/conftest.py
# conftest.py import os import random import boto3 import moto import pytest from moto.server import ThreadedMotoServer WD = os.getcwd() MODULE_PATH = "modules/analysis/rosbag-image-pipeline" if MODULE_PATH not in WD: WD = f"{WD}/{MODULE_PATH}" DAG_CONFIG_PATH = f"{WD}/image_dags/dag_config.py" DAG_CONFIG_BACKUP_PATH = f"{WD}/image_dags/dag_config.bak" def pytest_configure(config): """ Allows plugins and conftest files to perform initial configuration. This hook is called for every plugin and initial conftest file after command line options have been parsed. """ os.rename(DAG_CONFIG_PATH, DAG_CONFIG_BACKUP_PATH) sample_data = """SEEDFARMER_MODULE_METADATA = '{"PrivateSubnetIds":["subnet-090a22976151932d7","subnet-0d0f12bd07e5ed4ea","subnet-011bad0900787e44e"],"DagId":"vsi_image_pipeline","SecurityGroupId":"sg-08460867be55fd219","DagRoleArn":"arn:aws:iam::1234567890:role/addf-aws-solutions-analysis-rip-dag-us-east-1","DynamoDbTableName":"addf-aws-solutions-analysis-rip-drive-tracking","DetectionsDynamoDBName":"addf-aws-solutions-core-metadata-storage-Rosbag-Scene-Metadata","SourceBucketName":"addf-aws-solutions-raw-bucket-074ff5b4","TargetBucketName":"addf-aws-solutions-intermediate-bucket-074ff5b4","DagBucketName":"addf-aws-solutions-artifacts-bucket-074ff5b4","LogsBucketName":"addf-aws-solutions-logs-bucket-074ff5b4","OnDemandJobQueueArn":"arn:aws:batch:us-east-1:1234567890:job-queue/addf-aws-solutions-core-batch-compute-OnDemandJobQueue","SpotJobQueueArn":"arn:aws:batch:us-east-1:1234567890:job-queue/addf-aws-solutions-core-batch-compute-SpotJobQueue","FargateJobQueueArn":"arn:aws:batch:us-east-1:1234567890:job-queue/addf-aws-solutions-core-batch-compute-FargateJobQueue","ParquetBatchJobDefArn":"arn:aws:batch:us-east-1:1234567890:job-definition/addf-aws-solutions-docker-images-ros-to-parquet:1","PngBatchJobDefArn":"arn:aws:batch:us-east-1:1234567890:job-definition/addf-aws-solutions-docker-images-ros-to-png:1","ObjectDetectionImageUri":"1234567890.dkr.ecr.us-east-1.amazonaws.com/addf-aws-solutions-docker-images-object-detection:latest","ObjectDetectionRole":"arn:aws:iam::1234567890:role/addf-aws-solutions-docker-addfawssolutionsdockerim-1WI5F9LEEAN39","ObjectDetectionJobConcurrency":30,"ObjectDetectionInstanceType":"ml.m5.xlarge","LaneDetectionImageUri":"1234567890.dkr.ecr.us-east-1.amazonaws.com/addf-aws-solutions-docker-images-lane-detection:smprocessor","LaneDetectionRole":"arn:aws:iam::1234567890:role/addf-aws-solutions-docker-addfawssolutionsdockerim-1U2OPJ0QGMLSM","LaneDetectionJobConcurrency":20,"LaneDetectionInstanceType":"ml.m5.2xlarge","FileSuffix":".bag","DesiredEncoding":"bgr8","YoloModel":"yolov5s","ImageTopics":["/flir_adk/rgb_front_left/image_raw","/flir_adk/rgb_front_right/image_raw"],"SensorTopics":["/vehicle/gps/fix","/vehicle/gps/time","/vehicle/gps/vel","/imu_raw"]}' DEPLOYMENT_NAME = 'aws-solutions' MODULE_NAME = 'analysis-rip' REGION = 'us-east-1' EMR_JOB_EXECUTION_ROLE = 'arn:aws:iam::1234567890:role/addf-aws-solutions-core-e-addfawssolutionscoreemrs-KNO49U56R3MO' EMR_APPLICATION_ID = '00fcldsv5ol5dv09' SOLUTION_ID = 'SO12345' SOLUTION_NAME = 'Go ADDF!' SOLUTION_VERSION = 'v1.0.0' S3_SCRIPT_DIR = 's3://addf-aws-solutions-artifacts-bucket-074ff5b4/dags/aws-solutions/analysis-rip/image_dags/' """ # Writing to sample.json with open(DAG_CONFIG_PATH, "w") as outfile: outfile.write(sample_data) def pytest_unconfigure(config): """ called before test process is exited. """ os.remove(DAG_CONFIG_PATH) os.rename(DAG_CONFIG_BACKUP_PATH, DAG_CONFIG_PATH) @pytest.fixture(scope="function") def aws_credentials(): """Mocked AWS Credentials for moto.""" os.environ["AWS_ACCESS_KEY_ID"] = "testing" os.environ["AWS_SECRET_ACCESS_KEY"] = "testing" os.environ["AWS_SECURITY_TOKEN"] = "testing" os.environ["AWS_SESSION_TOKEN"] = "testing" os.environ["AWS_DEFAULT_REGION"] = "us-west-2" os.environ["MOTO_ACCOUNT_ID"] = "123456789012" @pytest.fixture(scope="function") def moto_dynamodb(aws_credentials): with moto.mock_dynamodb(): dynamodb = boto3.resource("dynamodb") dynamodb.create_table( TableName="mytable", KeySchema=[{"AttributeName": "pk", "KeyType": "HASH"}], AttributeDefinitions=[{"AttributeName": "pk", "AttributeType": "N"}], BillingMode="PAY_PER_REQUEST", ) yield dynamodb @pytest.fixture(scope="function") def moto_s3(aws_credentials): with moto.mock_s3(): s3 = boto3.client("s3") try: s3.create_bucket( Bucket="mybucket", CreateBucketConfiguration={"LocationConstraint": "us-west-2"}, ) except Exception as e: print(f"bucket creation failed: {e}") yield s3 @pytest.fixture(scope="function") def moto_server(): port = random.randint(5001, 8999) server = ThreadedMotoServer(port=port) server.start() yield port server.stop()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/__init__.py
modules/analysis/rosbag-image-pipeline/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/test_detect_scenes.py
modules/analysis/rosbag-image-pipeline/tests/test_detect_scenes.py
import os from image_dags.detect_scenes import * def create_spark_session(port: int): os.environ["PYSPARK_SUBMIT_ARGS"] = '--packages "org.apache.hadoop:hadoop-aws:3.3.1" pyspark-shell' spark = SparkSession.builder.getOrCreate() hadoop_conf = spark.sparkContext._jsc.hadoopConfiguration() hadoop_conf.set("fs.s3a.access.key", "dummy-value") hadoop_conf.set("fs.s3a.secret.key", "dummy-value") hadoop_conf.set("fs.s3a.endpoint", f"http://127.0.0.1:{port}") hadoop_conf.set("fs.s3.impl", "org.apache.hadoop.fs.s3a.S3AFileSystem") return spark def test_get_batch_file_metadata(moto_dynamodb): dynamodb = boto3.resource("dynamodb") table_name = "mytable" table = dynamodb.Table(table_name) items = [{"pk": 1}, {"pk": 2}] for item in items: table.put_item(Item=item) for item in items: result = get_batch_file_metadata(table_name, item["pk"], os.getenv("AWS_DEFAULT_REGION")) assert len(result) > 0 def test_detect_scenes_parse_arguments(): args = parse_arguments( [ "--batch-metadata-table-name", "dummy", "--batch-id", "dummy", "--input-bucket", "mybucket", "--output-bucket", "mybucket", "--output-dynamo-table", "mytable", "--region", "us-east-1", "--image-topics", '["/flir_adk/rgb_front_left/image_raw", "/flir_adk/rgb_front_right/image_raw"]', ] ) assert args.batch_metadata_table_name == "dummy" assert args.batch_id == "dummy" assert args.input_bucket == "mybucket" assert args.output_bucket == "mybucket" assert args.output_dynamo_table == "mytable" assert args.region == "us-east-1" assert args.image_topics == '["/flir_adk/rgb_front_left/image_raw", "/flir_adk/rgb_front_right/image_raw"]' def test_load_lane_detection(moto_server): port = moto_server s3 = boto3.resource("s3", endpoint_url=f"http://127.0.0.1:{port}") # create an S3 bucket. bucket_name = "lane-detection-bucket" s3.create_bucket( Bucket=bucket_name, CreateBucketConfiguration={"LocationConstraint": os.getenv("AWS_DEFAULT_REGION")}, ) object = s3.Object( bucket_name, "test-vehicle-02/small2__2020-11-19-16-21-22_4/_flir_adk_rgb_front_left_image_raw_resized_1280_720_post_lane_dets/lanes.csv", ) data = b"lanes" object.put(Body=data) spark = create_spark_session(port=port) sample_metadata = [ { "raw_image_bucket": bucket_name, "drive_id": "test-vehichle-01", "file_id": "this.jpg", "s3_bucket": bucket_name, "s3_key": "test-vehichle-01/this/_flir_adk_rgb_front_right_image_raw_resized_/1280_720_post_lane_dets/lanes.csv", "resized_image_dirs": [ "test-vehicle-02/small2__2020-11-19-16-21-22_4/_flir_adk_rgb_front_left_image_raw_resized_1280_720", "test-vehicle-02/small2__2020-11-19-16-21-22_4/_flir_adk_rgb_front_right_image_raw_resized_1280_720", ], } ] load_lane_detection(spark, sample_metadata) spark.stop() def test_load_obj_detection(moto_server): port = moto_server s3 = boto3.resource("s3", endpoint_url=f"http://127.0.0.1:{port}") s3.create_bucket( Bucket="mybucket2", CreateBucketConfiguration={"LocationConstraint": os.getenv("AWS_DEFAULT_REGION")}, ) object = s3.Object( "mybucket2", "test-vehicle-02/small2__2020-11-19-16-21-22_4/_flir_adk_rgb_front_left_image_raw_resized_1280_720_post_obj_dets/all_predictions.csv", ) data = b"all_predictions" object.put(Body=data) spark = create_spark_session(port=port) sample_metadata = [ { "raw_image_bucket": "mybucket2", "drive_id": "test-vehichle-01", "file_id": "this.jpg", "resized_image_dirs": [ "test-vehicle-02/small2__2020-11-19-16-21-22_4/_flir_adk_rgb_front_left_image_raw_resized_1280_720", ], } ] load_obj_detection(spark, sample_metadata, None) spark.stop() def test_write_results_to_s3(moto_server): port = moto_server s3 = boto3.resource("s3", endpoint_url=f"http://127.0.0.1:{port}") s3.create_bucket( Bucket="outputbucket", CreateBucketConfiguration={"LocationConstraint": os.getenv("AWS_DEFAULT_REGION")}, ) spark = create_spark_session(port=port) df = spark.createDataFrame( [ (1, "foo"), (2, "bar"), ], ["id", "bag_file"], ) dfs = {"test": df} write_results_s3( dfs, table_name="scene_detections", output_bucket="outputbucket", partition_cols=["bag_file"], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/test_batch_creation_and_tracking.py
modules/analysis/rosbag-image-pipeline/tests/test_batch_creation_and_tracking.py
from image_dags.batch_creation_and_tracking import * def test_get_batch_file_metadata(moto_dynamodb): drives_and_files = {"foo": {"files": ["dummyfile"], "bucket": "dummybucket"}} batch_write_files_to_dynamo( table=moto_dynamodb.Table("mytable"), drives_and_files=drives_and_files, batch_id=1010, ) def test_get_drive_files(moto_s3): get_drive_files( src_bucket="mybucket", src_prefix="files/", file_suffix=".bag", s3_client=moto_s3, ) def test_add_drives_to_batch(moto_dynamodb, moto_s3): table = moto_dynamodb.Table("mytable") drives_to_process = { "drive1": {"bucket": "mybucket", "prefix": "rosbag-scene-detection/drive1/"}, "drive2": {"bucket": "mybucket", "prefix": "rosbag-scene-detection/drive2/"}, } add_drives_to_batch( table=table, batch_id=1010, drives_to_process=drives_to_process, file_suffix=".bag", s3_client=moto_s3, )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/tests/test_ros_image_pipeline.py
modules/analysis/rosbag-image-pipeline/tests/test_ros_image_pipeline.py
from image_dags.ros_image_pipeline import * def test_validate_config(): input = { "drive2": { "bucket": "addf-ros-image-demo-raw-bucket-d2be7d29", "prefix": "rosbag-scene-detection/drive2/", }, } validate_config(input) def test_get_job_name(): job_name = get_job_name("foobar") assert job_name.startswith("ros-image-pipeline-foobar-")
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/image_dags/dag_config.py
modules/analysis/rosbag-image-pipeline/image_dags/dag_config.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 # This file is populated with configurations information when the Module is deployed # Configuration parameters are exported as module level constants # # Example: # SOME_PARAMETER = 'some value' # DUMPS PARAMETERS FROM app.py IN ../deployspec.yaml like: # - echo "DEPLOYMENT_NAME = '${SEEDFARMER_DEPLOYMENT_NAME}'" >> image_dags/dag_config.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/image_dags/detect_scenes.py
modules/analysis/rosbag-image-pipeline/image_dags/detect_scenes.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import argparse import json import sys from typing import Any, Dict, List import boto3 import pyspark.sql.functions as func from pyspark.sql import SparkSession, Window from pyspark.sql.functions import ( aggregate, col, concat, count, first, from_json, lit, split, ) from pyspark.sql.types import ArrayType, DoubleType, IntegerType, StringType, StructField, StructType obj_schema = StructType( [ StructField("_c0", IntegerType(), True), StructField("xmin", DoubleType(), True), StructField("ymin", DoubleType(), True), StructField("xmax", DoubleType(), True), StructField("ymax", DoubleType(), True), StructField("confidence", DoubleType(), True), StructField("class", IntegerType(), True), StructField("name", StringType(), True), StructField("source_image", StringType(), True), ] ) def parse_arguments(args): parser = argparse.ArgumentParser() parser.add_argument("--batch-metadata-table-name", required=True) parser.add_argument("--batch-id", required=True) parser.add_argument("--input-bucket", required=False) parser.add_argument("--output-bucket", required=True) parser.add_argument("--output-dynamo-table", required=True) parser.add_argument("--region", required=True) parser.add_argument("--image-topics", required=False) return parser.parse_args(args=args) def form_object_in_lane_df(obj_df, lane_df): obj_lane_df = obj_df.join(lane_df, on="source_image", how="inner") obj_lane_df = ( obj_lane_df.withColumn( "pixel_rows_at_bottom_corner", obj_lane_df["lanes"].getItem(obj_lane_df.ymax.cast(IntegerType())), ) .drop("lanes") .drop("ymax") ) obj_lane_df = obj_lane_df.withColumn( "sum_of_pixels_intensities_at_bottom_corner_rows", aggregate("pixel_rows_at_bottom_corner", lit(0), lambda acc, x: acc + x), ).drop("pixel_rows_at_bottom_corner") obj_lane_df = obj_lane_df.filter(obj_lane_df.sum_of_pixels_intensities_at_bottom_corner_rows != 0) obj_lane_df = obj_lane_df.withColumn("source_image_split", split("source_image", "_")) obj_lane_df = obj_lane_df.withColumn( "Time", concat( obj_lane_df["source_image_split"].getItem(2), lit("."), obj_lane_df["source_image_split"].getItem(3).substr(1, 2), ).cast(DoubleType()), ) return obj_lane_df def get_batch_file_metadata(table_name, batch_id, region): dynamodb = boto3.resource("dynamodb", region_name=region) table = dynamodb.Table(table_name) response = table.query(KeyConditions={"pk": {"AttributeValueList": [batch_id], "ComparisonOperator": "EQ"}}) data = response["Items"] while "LastEvaluatedKey" in response: response = table.query(ExclusiveStartKey=response["LastEvaluatedKey"]) data.update(response["Items"]) return data def load_obj_detection(spark, batch_metadata, image_topics): path_list = [] def _process( image_topics: List[str], path_list: List[str], item: Dict[str, Any], resizied_image_dir: str, ) -> None: for t in image_topics: path_list.append( f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_obj_dets/all_predictions.csv" ) if t in resizied_image_dir else None for item in batch_metadata: for resizied_image_dir in item["resized_image_dirs"]: if image_topics and len(image_topics) > 0: _process(image_topics, path_list, item, resizied_image_dir) else: path_list.append( f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_obj_dets/all_predictions.csv" ) def remove_space(name): return name.replace(" ", "_") remove_space_udf = func.udf(remove_space, StringType()) df = spark.read.schema(obj_schema).option("header", True).csv(path_list) df2 = df.withColumn("name", remove_space_udf(df.name)) print("Number of rows in Object Detection dataframe") print(df2.count()) return df2 def load_lane_detection(spark, batch_metadata): first_item = batch_metadata[0] first_path_prefix = first_item["resized_image_dirs"][0] first_path = f"s3://{first_item['raw_image_bucket']}/{first_path_prefix}_post_lane_dets/lanes.csv" first_item_split = first_item["s3_key"].rpartition("/") bag_file_prefix = first_item_split[0] bag_file = first_item_split[2].split(".")[0] df = ( spark.read.option("header", True) .csv(first_path) .withColumn("bag_file_bucket", lit(first_item["s3_bucket"])) .withColumn("bag_file", lit(bag_file)) .withColumn("bag_file_prefix", lit(bag_file_prefix)) ) for i in range(1, len(batch_metadata)): item = batch_metadata[i] for resizied_image_dir in item["resized_image_dirs"]: path = f"s3://{item['raw_image_bucket']}/{resizied_image_dir}_post_lane_dets/lanes.csv" item_split = item["s3_key"].rpartition("/") bag_file_prefix = item_split[0] bag_file = item_split[2].split(".")[0] df.union( spark.read.option("header", True) .csv(path) .withColumn("bag_file_bucket", lit(item["s3_bucket"])) .withColumn("bag_file", lit(bag_file)) .withColumn("bag_file_prefix", lit(bag_file_prefix)) ) lane_schema = ArrayType(ArrayType(IntegerType()), False) df = df.withColumn("lanes", from_json(col("lanes"), lane_schema)) return df def write_results_s3(dfs: Dict[str, Any], table_name: str, output_bucket: str, partition_cols=[]): for key, df in dfs.items(): s3_path = f"s3://{output_bucket}/{table_name}/{key}" df.write.mode("append").partitionBy(*partition_cols).parquet(s3_path) def write_results_dynamo(dfs, output_dynamo_table, region): for key, df in dfs.items(): print(f"Writing {key} to DDB") df.write.mode("append").option("tableName", output_dynamo_table).option("region", region).format( "dynamodb" ).save() def summarize_obj_in_lane_scenes(obj_lane_df, image_topics, obj_type): obj_lane_df = obj_lane_df.filter(obj_lane_df.name == obj_type) obj_lane_df = obj_lane_df.groupby("source_image").agg( count("name").alias(f"num_{obj_type}_in_lane"), first("Time").alias("Time"), first("bag_file_bucket").alias("bag_file_bucket"), first("bag_file").alias("bag_file"), first("bag_file_prefix").alias("bag_file_prefix"), ) win = Window.orderBy("Time").partitionBy("bag_file_bucket", "bag_file", "bag_file_prefix") obj_lane_df = ( obj_lane_df.withColumn( f"num_{obj_type}_in_lane_lag1", func.lag(func.col(f"num_{obj_type}_in_lane"), 1, 0).over(win), ) .withColumn( f"num_{obj_type}_in_lane_lead1", func.lead(func.col(f"num_{obj_type}_in_lane"), 1, 0).over(win), ) .filter(f"num_{obj_type}_in_lane_lag1 == 0 or num_{obj_type}_in_lane_lead1 ==0") ) scene_state_udf = func.udf( lambda num, lag: "start" if num > 0 and lag is None else ("end" if num == 0 and lag > 0 else None), StringType(), ) obj_in_lane_scenes_df = ( obj_lane_df.withColumn( "scene_state", scene_state_udf( obj_lane_df[f"num_{obj_type}_in_lane"], obj_lane_df[f"num_{obj_type}_in_lane_lag1"], ), ) .withColumn("end_time", func.lead(func.col("Time"), 1).over(win)) .filter(f"num_{obj_type}_in_lane_lag1 ==0") .withColumnRenamed("Time", "start_time") .withColumnRenamed(f"num_{obj_type}_in_lane", f"num_{obj_type}_in_lane_start") .select( "bag_file", "bag_file_bucket", "bag_file_prefix", "start_time", "end_time", f"num_{obj_type}_in_lane_start", ) .withColumn( "scene_id", func.concat( func.col("bag_file"), func.lit(f"_{obj_type}InLane_"), func.col("start_time"), ), ) .withColumn("scene_length", func.col("end_time") - func.col("start_time")) .withColumn("topics_analyzed", func.lit(",".join(image_topics))) ) return obj_in_lane_scenes_df def main( batch_metadata_table_name, batch_id, output_bucket, output_dynamo_table, spark, region, image_topics, ): # Load files to process batch_metadata = get_batch_file_metadata(table_name=batch_metadata_table_name, batch_id=batch_id, region=region) if image_topics: image_topics = json.loads(image_topics) image_topics = [topic.replace("/", "_") for topic in image_topics if image_topics] # Load topic data from s3 and union obj_df = load_obj_detection(spark, batch_metadata=batch_metadata, image_topics=image_topics) lane_df = load_lane_detection( spark, batch_metadata=batch_metadata, ) obj_lane_df = form_object_in_lane_df(obj_df, lane_df) dfs = {} dfs["car"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "car") dfs["truck"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "truck") dfs["traffic_light"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "traffic_light") dfs["train"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "train") dfs["bus"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "bus") dfs["motorcycle"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "motorcycle") dfs["stop_sign"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "stop_sign") dfs["fire_hydrant"] = summarize_obj_in_lane_scenes(obj_lane_df, image_topics, "fire_hydrant") write_results_s3( dfs, table_name="scene_detections", output_bucket=output_bucket, partition_cols=["bag_file"], ) write_results_dynamo(dfs, output_dynamo_table, region) if __name__ == "__main__": spark = SparkSession.builder.appName("scene-detection").getOrCreate() sc = spark.sparkContext arguments = parse_arguments(sys.argv[1:]) batch_metadata_table_name = arguments.batch_metadata_table_name batch_id = arguments.batch_id output_bucket = arguments.output_bucket output_dynamo_table = arguments.output_dynamo_table region = arguments.region image_topics = arguments.image_topics if arguments.image_topics else None main( batch_metadata_table_name, batch_id, output_bucket, output_dynamo_table, spark, region, image_topics, ) sc.stop()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/image_dags/batch_creation_and_tracking.py
modules/analysis/rosbag-image-pipeline/image_dags/batch_creation_and_tracking.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging import typing logger = logging.getLogger("airflow") logger.setLevel("DEBUG") def add_drives_to_batch( table: str, batch_id: str, drives_to_process: typing.Dict[str, dict], file_suffix: str, s3_client, ): """Lists files with file_suffix for each prefix in drives_to_process and adds each file to dynamodb for tracking @param table: dynamo tracking table @param batch_id: dag run id @param drives_to_process: { "drives_to_process": { "drive1": {"bucket": "addf-example-dev-raw-bucket-xyz", "prefix": "rosbag-scene-detection/drive1/"}, "drive2": {"bucket": "addf-example-dev-raw-bucket-xyz", "prefix": "rosbag-scene-detection/drive2/"}, }, } @param file_suffix: ".bag" @param s3_client: type boto3.client('s3') @return: """ drives_and_files = {} files_in_batch = 0 for drive_id, s3_path in drives_to_process.items(): files = get_drive_files( src_bucket=s3_path["bucket"], src_prefix=s3_path["prefix"], file_suffix=file_suffix, s3_client=s3_client, ) drives_and_files[drive_id] = {"files": files, "bucket": s3_path["bucket"]} files_in_batch += len(files) logger.info(f"files_in_batch {files_in_batch}") batch_write_files_to_dynamo(table, drives_and_files, batch_id) return files_in_batch def get_drive_files(src_bucket, src_prefix, file_suffix, s3_client): """For a given bucket, prefix, and suffix, lists all files found on S3 and returns a list of the files @param drive_id: @param src_bucket: @param src_prefix: @param file_suffix: @param s3_client: @return: """ MAX_KEYS = 1000 logger.info(src_bucket) logger.info(src_prefix) file_response = s3_client.list_objects_v2(Bucket=src_bucket, Prefix=src_prefix, MaxKeys=MAX_KEYS, Delimiter="/") logger.info(file_response) file_next_continuation = file_response.get("NextContinuationToken") files = [x["Key"] for x in file_response.get("Contents", []) if x["Key"].endswith(file_suffix)] while file_next_continuation is not None: file_response = s3_client.list_objects_v2( Bucket=src_bucket, Prefix=src_prefix, MaxKeys=MAX_KEYS, Delimiter="/", ContinuationToken=file_next_continuation, ) file_next_continuation = file_response.get("NextContinuationToken") files += [x["Key"] for x in file_response.get("Contents", [])] logger.info(files) return files def batch_write_files_to_dynamo(table, drives_and_files, batch_id): with table.batch_writer() as batch: idx = 0 for drive_id, files in drives_and_files.items(): for file in files["files"]: item = { "drive_id": drive_id, "file_id": file.split("/")[-1], "s3_bucket": files["bucket"], "s3_key": file, "pk": batch_id, "sk": str(idx), } logger.info(item) batch.put_item(Item=item) idx += 1
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/image_dags/ros_image_pipeline.py
modules/analysis/rosbag-image-pipeline/image_dags/ros_image_pipeline.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import json import logging import os import random import string import sys import time from datetime import timedelta from math import ceil from typing import TypeVar import boto3 import botocore from airflow import DAG, settings from airflow.contrib.hooks.aws_hook import AwsHook from airflow.exceptions import AirflowException from airflow.models import Connection from airflow.operators.python import PythonOperator, get_current_context from airflow.providers.amazon.aws.operators.batch import BatchOperator from airflow.providers.amazon.aws.operators.emr import EmrServerlessStartJobOperator from airflow.providers.amazon.aws.sensors.emr import EmrServerlessJobSensor from airflow.utils.dates import days_ago from airflow.utils.task_group import TaskGroup from boto3.dynamodb.conditions import Key from sagemaker.network import NetworkConfig from sagemaker.processing import ProcessingInput, ProcessingOutput, Processor sys.path.insert(0, os.path.abspath(os.path.dirname(__file__))) from batch_creation_and_tracking import add_drives_to_batch from dag_config import ( EMR_APPLICATION_ID, EMR_JOB_EXECUTION_ROLE, REGION, S3_SCRIPT_DIR, SEEDFARMER_MODULE_METADATA, SOLUTION_ID, SOLUTION_VERSION, ) # GET MODULE VARIABLES FROM APP.PY AND DEPLOYSPEC module_metadata = json.loads(SEEDFARMER_MODULE_METADATA) DAG_ID = module_metadata["DagId"] DAG_ROLE = module_metadata["DagRoleArn"] DYNAMODB_TABLE = module_metadata["DynamoDbTableName"] FARGATE_JOB_QUEUE_ARN = module_metadata["FargateJobQueueArn"] ON_DEMAND_JOB_QUEUE_ARN = module_metadata["OnDemandJobQueueArn"] SPOT_JOB_QUEUE_ARN = module_metadata["SpotJobQueueArn"] TARGET_BUCKET = module_metadata["TargetBucketName"] FILE_SUFFIX = module_metadata["FileSuffix"] PRIVATE_SUBNETS_IDS = module_metadata["PrivateSubnetIds"] SM_SECURITY_GROUP_ID = module_metadata["SecurityGroupId"] PNG_JOB_DEFINITION_ARN = module_metadata["PngBatchJobDefArn"] DESIRED_ENCODING = module_metadata["DesiredEncoding"] IMAGE_TOPICS = module_metadata["ImageTopics"] PARQUET_JOB_DEFINITION_ARN = module_metadata["ParquetBatchJobDefArn"] SENSOR_TOPICS = module_metadata["SensorTopics"] YOLO_IMAGE_URI = module_metadata["ObjectDetectionImageUri"] YOLO_ROLE = module_metadata["ObjectDetectionRole"] YOLO_CONCURRENCY = module_metadata["ObjectDetectionJobConcurrency"] YOLO_INSTANCE_TYPE = module_metadata["ObjectDetectionInstanceType"] YOLO_MODEL = module_metadata["YoloModel"] LANEDET_IMAGE_URI = module_metadata["LaneDetectionImageUri"] LANEDET_ROLE = module_metadata["LaneDetectionRole"] LANEDET_CONCURRENCY = module_metadata["LaneDetectionJobConcurrency"] LANEDET_INSTANCE_TYPE = module_metadata["LaneDetectionInstanceType"] # EMR Config # spark_app_dir = f"s3://{module_metadata['DagBucketName']}/spark_jobs/" # EMR_VIRTUAL_CLUSTER_ID = module_metadata['EmrVirtualClusterId'] # EMR_JOB_ROLE_ARN = module_metadata['EmrJobRoleArn'] # ARTIFACT_BUCKET = module_metadata["DagBucketName"] LOGS_BUCKET = module_metadata["LogsBucketName"] SCENE_TABLE = module_metadata["DetectionsDynamoDBName"] CONFIGURATION_OVERRIDES = { "monitoringConfiguration": { "managedPersistenceMonitoringConfiguration": {"enabled": True}, "s3MonitoringConfiguration": {"logUri": f"s3://{LOGS_BUCKET}/scene-detection"}, } } ValueType = TypeVar("ValueType") TASK_DEF_XCOM_KEY = "job_definition_arn" DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("DEBUG") def get_config() -> botocore.config.Config: if SOLUTION_ID and SOLUTION_VERSION: user_agent_extra_param = {"user_agent_extra": f"AWSSOLUTION/{SOLUTION_ID}/{SOLUTION_VERSION}"} return botocore.config.Config(**user_agent_extra_param) else: return botocore.config.Config() def try_create_aws_conn(): conn_id = "aws_default" try: AwsHook.get_connection(conn_id) except AirflowException: extra = json.dumps({"role_arn": DAG_ROLE}, indent=2) conn = Connection(conn_id=conn_id, conn_type="aws", host="", schema="", login="", extra=extra) try: session = settings.Session() session.add(conn) session.commit() finally: session.close() def validate_config(drives_to_process): example_input = { "drives_to_process": { "drive2": { "bucket": "addf-ros-image-demo-raw-bucket-d2be7d29", "prefix": "rosbag-scene-detection/drive2/", }, } } for k, v in drives_to_process.items(): assert isinstance(k, str), f"expecting config to be like {example_input}, received: {drives_to_process}" assert ( "bucket" in v.keys() and "prefix" in v.keys() ), f"expecting config to be like {example_input}, received: {drives_to_process}" assert v["prefix"][-1] == "/" def create_batch_of_drives(ti, **kwargs): """ if Batch Id already exists, then run dag again on exact same files Else: List Drive Folders in S3 For New Drives, List Recording Files Put New Drives and Recording Files in Dynamo and Assign to a batch if not assigned already Add Drives until reaching max files allowed in 1 batch (hard limit of 10k) """ drives_to_process = kwargs["dag_run"].conf["drives_to_process"] batch_id = kwargs["dag_run"].run_id # Establish AWS API Connections sts_client = boto3.client("sts") assumed_role_object = sts_client.assume_role(RoleArn=DAG_ROLE, RoleSessionName="AssumeRoleSession1") credentials = assumed_role_object["Credentials"] dynamodb = boto3.resource( "dynamodb", aws_access_key_id=credentials["AccessKeyId"], aws_secret_access_key=credentials["SecretAccessKey"], aws_session_token=credentials["SessionToken"], config=get_config(), ) s3_client = boto3.client( "s3", aws_access_key_id=credentials["AccessKeyId"], aws_secret_access_key=credentials["SecretAccessKey"], aws_session_token=credentials["SessionToken"], config=get_config(), ) # Validate Config validate_config(drives_to_process) table = dynamodb.Table(DYNAMODB_TABLE) files_in_batch = table.query( KeyConditionExpression=Key("pk").eq(batch_id), Select="COUNT", )["Count"] if files_in_batch > 0: logger.info("Batch Id already exists in tracking table - using existing batch") return files_in_batch logger.info("New Batch Id - collecting unprocessed drives from S3 and adding to the batch") files_in_batch = add_drives_to_batch( table=table, drives_to_process=drives_to_process, batch_id=batch_id, file_suffix=FILE_SUFFIX, s3_client=s3_client, ) assert files_in_batch <= 10000, "AWS Batch Array Size cannot exceed 10000" return files_in_batch def get_job_name(suffix="") -> str: v = "".join(random.choice(string.ascii_lowercase) for _i in range(6)) return f"ros-image-pipeline-{suffix}-{v}" def png_batch_operation(**kwargs): logger.info(f"kwargs at png_batch_operations is {kwargs}") ti = kwargs["ti"] array_size = ti.xcom_pull(task_ids="create-batch-of-drives", key="return_value") batch_id = kwargs["dag_run"].run_id context = get_current_context() op = BatchOperator( task_id="submit_batch_job_op", job_name=get_job_name("png"), job_queue=ON_DEMAND_JOB_QUEUE_ARN, aws_conn_id="aws_default", job_definition=PNG_JOB_DEFINITION_ARN, array_properties={"size": int(array_size)}, overrides={ "environment": [ {"name": "TABLE_NAME", "value": DYNAMODB_TABLE}, {"name": "BATCH_ID", "value": batch_id}, {"name": "DEBUG", "value": "true"}, {"name": "IMAGE_TOPICS", "value": json.dumps(IMAGE_TOPICS)}, {"name": "DESIRED_ENCODING", "value": DESIRED_ENCODING}, {"name": "TARGET_BUCKET", "value": TARGET_BUCKET}, ], }, ) op.execute(context) def parquet_operation(**kwargs): ti = kwargs["ti"] array_size = ti.xcom_pull(task_ids="create-batch-of-drives", key="return_value") batch_id = kwargs["dag_run"].run_id context = get_current_context() op = BatchOperator( task_id="submit_parquet_job_op", job_name=get_job_name("parq"), job_queue=FARGATE_JOB_QUEUE_ARN, aws_conn_id="aws_default", job_definition=PARQUET_JOB_DEFINITION_ARN, array_properties={"size": int(array_size)}, overrides={ "environment": [ {"name": "TABLE_NAME", "value": DYNAMODB_TABLE}, {"name": "BATCH_ID", "value": batch_id}, {"name": "TOPICS", "value": json.dumps(SENSOR_TOPICS)}, {"name": "TARGET_BUCKET", "value": TARGET_BUCKET}, ], }, ) op.execute(context) def sagemaker_yolo_operation(**kwargs): # Establish AWS API Connections sts_client = boto3.client("sts") assumed_role_object = sts_client.assume_role(RoleArn=DAG_ROLE, RoleSessionName="AssumeRoleSession1") credentials = assumed_role_object["Credentials"] dynamodb = boto3.resource( "dynamodb", aws_access_key_id=credentials["AccessKeyId"], aws_secret_access_key=credentials["SecretAccessKey"], aws_session_token=credentials["SessionToken"], config=get_config(), ) table = dynamodb.Table(DYNAMODB_TABLE) batch_id = kwargs["dag_run"].run_id # Get Image Directories per Recording File to Label image_directory_items = table.query( KeyConditionExpression=Key("pk").eq(batch_id), Select="SPECIFIC_ATTRIBUTES", ProjectionExpression="resized_image_dirs", )["Items"] image_directories = [] for item in image_directory_items: image_directories += item["resized_image_dirs"] logger.info(f"Starting object detection job for {len(image_directories)} directories") total_jobs = len(image_directories) num_batches = ceil(total_jobs / YOLO_CONCURRENCY) for i in range(num_batches): logger.info(f"Starting object detection job for batch {i + 1} of {num_batches}") processor = Processor( image_uri=YOLO_IMAGE_URI, role=YOLO_ROLE, instance_count=1, instance_type=YOLO_INSTANCE_TYPE, base_job_name=f"{batch_id.replace(':', '').replace('_', '')[0:23]}-YOLO", network_config=NetworkConfig(subnets=PRIVATE_SUBNETS_IDS, security_group_ids=[SM_SECURITY_GROUP_ID]), ) idx_start = i * YOLO_CONCURRENCY idx_end = (i + 1) * YOLO_CONCURRENCY for image_directory in image_directories[idx_start:idx_end]: logger.info(f"Starting object detection job for {image_directory}") logger.info( "Job details available at: " f"https://{REGION}.console.aws.amazon.com/sagemaker/home?region={REGION}#/processing-jobs" ) processor.run( inputs=[ ProcessingInput( input_name="data", source=f"s3://{TARGET_BUCKET}/{image_directory}/", destination="/opt/ml/processing/input/", ) ], outputs=[ ProcessingOutput( output_name="output", source="/opt/ml/processing/output/", destination=f"s3://{TARGET_BUCKET}/{image_directory}_post_obj_dets/", ) ], arguments=["--model", YOLO_MODEL], wait=False, logs=False, ) time.sleep(1) # Attempt at avoiding throttling exceptions logger.info("Waiting on batch of jobs to finish") for job in processor.jobs: logger.info(f"Waiting on: {job} - logs from job:") job.wait(logs=False) logger.info("All object detection jobs complete") def sagemaker_lanedet_operation(**kwargs): # Establish AWS API Connections sts_client = boto3.client("sts") assumed_role_object = sts_client.assume_role(RoleArn=DAG_ROLE, RoleSessionName="AssumeRoleSession1") credentials = assumed_role_object["Credentials"] dynamodb = boto3.resource( "dynamodb", aws_access_key_id=credentials["AccessKeyId"], aws_secret_access_key=credentials["SecretAccessKey"], aws_session_token=credentials["SessionToken"], config=get_config(), ) table = dynamodb.Table(DYNAMODB_TABLE) batch_id = kwargs["dag_run"].run_id # Get Image Directories per Recording File to Label image_directory_items = table.query( KeyConditionExpression=Key("pk").eq(batch_id), Select="SPECIFIC_ATTRIBUTES", ProjectionExpression="resized_image_dirs", )["Items"] image_directories = [] for item in image_directory_items: image_directories += item["resized_image_dirs"] logger.info(f"Starting lane detection job for {len(image_directories)} directories") total_jobs = len(image_directories) num_batches = ceil(total_jobs / LANEDET_CONCURRENCY) for i in range(num_batches): logger.info(f"Starting lane detection job for batch {i + 1} of {num_batches}") processor = Processor( image_uri=LANEDET_IMAGE_URI, role=LANEDET_ROLE, instance_count=1, instance_type=LANEDET_INSTANCE_TYPE, base_job_name=f"{batch_id.replace(':', '').replace('_', '')[0:23]}-LANE", network_config=NetworkConfig(subnets=PRIVATE_SUBNETS_IDS, security_group_ids=[SM_SECURITY_GROUP_ID]), ) LOCAL_INPUT = "/opt/ml/processing/input/image" LOCAL_OUTPUT = "/opt/ml/processing/output/image" LOCAL_OUTPUT_JSON = "/opt/ml/processing/output/json" LOCAL_OUTPUT_CSV = "/opt/ml/processing/output/csv" idx_start = i * LANEDET_CONCURRENCY idx_end = (i + 1) * LANEDET_CONCURRENCY for image_directory in image_directories[idx_start:idx_end]: logger.info(f"Starting lane detection job for {image_directory}") logger.info( "Job details available at: " f"https://{REGION}.console.aws.amazon.com/sagemaker/home?region={REGION}#/processing-jobs" ) processor.run( arguments=[ "--save_dir", LOCAL_OUTPUT, "--source", LOCAL_INPUT, "--json_path", LOCAL_OUTPUT_JSON, "--csv_path", LOCAL_OUTPUT_CSV, ], inputs=[ ProcessingInput( input_name="data", source=f"s3://{TARGET_BUCKET}/{image_directory}/", destination=LOCAL_INPUT, ) ], outputs=[ ProcessingOutput( output_name="image_output", source=LOCAL_OUTPUT, destination=f"s3://{TARGET_BUCKET}/{image_directory}_post_lane_dets/", ), ProcessingOutput( output_name="json_output", source=LOCAL_OUTPUT_JSON, destination=f"s3://{TARGET_BUCKET}/{image_directory}_post_lane_dets/", ), ProcessingOutput( output_name="csv_output", source=LOCAL_OUTPUT_CSV, destination=f"s3://{TARGET_BUCKET}/{image_directory}_post_lane_dets/", ), ], wait=False, logs=False, ) time.sleep(1) # Attempt at avoiding throttling exceptions logger.info("Waiting on batch of jobs to finish") for job in processor.jobs: logger.info(f"Waiting on: {job} - logs from job:") job.wait(logs=False) logger.info("All object detection jobs complete") def emr_batch_operation(**kwargs): ds = kwargs["ds"] batch_id = kwargs["dag_run"].run_id JOB_DRIVER = { "sparkSubmit": { "entryPoint": f"{S3_SCRIPT_DIR}detect_scenes.py", "entryPointArguments": [ "--batch-metadata-table-name", DYNAMODB_TABLE, "--batch-id", batch_id, "--output-bucket", TARGET_BUCKET, "--region", REGION, "--output-dynamo-table", SCENE_TABLE, "--image-topics", json.dumps(IMAGE_TOPICS), ], "sparkSubmitParameters": f"--jars {S3_SCRIPT_DIR}spark-dynamodb_2.12-1.1.1.jar", } } start_job_run_op = EmrServerlessStartJobOperator( task_id="scene_detection", application_id=EMR_APPLICATION_ID, execution_role_arn=EMR_JOB_EXECUTION_ROLE, job_driver=JOB_DRIVER, configuration_overrides=CONFIGURATION_OVERRIDES, aws_conn_id="aws_default", ) job_run_id = start_job_run_op.execute(ds) return job_run_id with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=days_ago(1), # type: ignore schedule_interval="@once", render_template_as_native_obj=True, ) as dag: create_aws_conn = PythonOperator( task_id="try-create-aws-conn", python_callable=try_create_aws_conn, dag=dag, ) create_batch_of_drives_task = PythonOperator( task_id="create-batch-of-drives", python_callable=create_batch_of_drives, dag=dag, provide_context=True, ) # Start Task Group definition with TaskGroup(group_id="sensor-extraction") as extract_task_group: submit_png_job = PythonOperator(task_id="image-extraction-batch-job", python_callable=png_batch_operation) submit_parquet_job = PythonOperator(task_id="parquet-extraction-batch-job", python_callable=parquet_operation) create_batch_of_drives_task >> [submit_parquet_job, submit_png_job] with TaskGroup(group_id="image-labelling") as image_labelling_task_group: submit_yolo_job = PythonOperator( task_id="object-detection-sagemaker-job", python_callable=sagemaker_yolo_operation, ) submit_lane_det_job = PythonOperator( task_id="lane-detection-sagemaker-job", python_callable=sagemaker_lanedet_operation, ) with TaskGroup(group_id="scene-detection") as scene_detection_task_group: start_job_run = PythonOperator(task_id="scene-detection", python_callable=emr_batch_operation) job_sensor = EmrServerlessJobSensor( task_id="check-emr-job-status", application_id=EMR_APPLICATION_ID, job_run_id="{{ task_instance.xcom_pull(task_ids='scene-detection.scene-detection', key='return_value') }}", aws_conn_id="aws_default", ) start_job_run >> job_sensor create_aws_conn >> create_batch_of_drives_task >> extract_task_group submit_png_job >> image_labelling_task_group >> scene_detection_task_group
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/analysis/rosbag-image-pipeline/image_dags/__init__.py
modules/analysis/rosbag-image-pipeline/image_dags/__init__.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License.
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/optionals/datalake-buckets/stack.py
modules/optionals/datalake-buckets/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import hashlib import logging from typing import Any, cast import aws_cdk import aws_cdk.aws_iam as aws_iam import aws_cdk.aws_s3 as aws_s3 import cdk_nag from aws_cdk import Aspects, Duration, Stack, Tags from cdk_nag import NagPackSuppression, NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class DataLakeBucketsStack(Stack): def __init__( self, scope: Construct, id: str, project_name: str, deployment_name: str, module_name: str, hash: str, buckets_encryption_type: str, buckets_retention: str, artifacts_log_retention: int, stack_description: str, **kwargs: Any, ) -> None: # CDK Env Vars account: str = aws_cdk.Aws.ACCOUNT_ID region: str = aws_cdk.Aws.REGION partition: str = aws_cdk.Aws.PARTITION dep_mod = f"{project_name}-{deployment_name}-{module_name}" # used to tag AWS resources. Tag Value length cant exceed 256 characters full_dep_mod = dep_mod[:256] if len(dep_mod) > 256 else dep_mod super().__init__(scope, id, description=stack_description, **kwargs) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=full_dep_mod) logs_bucket_name = f"{project_name}-{deployment_name}-logs-bucket-{hash}" hashlib.sha256() unique_ab = (hashlib.sha256(module_name.encode("utf-8")).hexdigest())[: (60 - len(logs_bucket_name))] logs_bucket = aws_s3.Bucket( self, id="logs-bucket", bucket_name=f"{logs_bucket_name}-{unique_ab}", removal_policy=aws_cdk.RemovalPolicy.RETAIN if buckets_retention.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, auto_delete_objects=None if buckets_retention.upper() == "RETAIN" else True, # Encryption should be always set to AES256 for a bucket to receive access logging from target buckets encryption=aws_s3.BucketEncryption.S3_MANAGED, block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, object_ownership=aws_s3.ObjectOwnership.BUCKET_OWNER_PREFERRED, enforce_ssl=True, # MWAA is very chatty, logs need to be cleaned via LifecycleRule lifecycle_rules=[ aws_s3.LifecycleRule( expiration=Duration.days(artifacts_log_retention), enabled=True, prefix="artifacts-bucket-logs/", ) ], ) raw_bucket_name = f"{project_name}-{deployment_name}-raw-bucket-{hash}" unique_ab = (hashlib.sha256(module_name.encode("utf-8")).hexdigest())[: (60 - len(raw_bucket_name))] raw_bucket = aws_s3.Bucket( self, removal_policy=aws_cdk.RemovalPolicy.RETAIN if buckets_retention.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, bucket_name=f"{raw_bucket_name}-{unique_ab}", versioned=True, server_access_logs_bucket=logs_bucket, server_access_logs_prefix="raw-bucket-logs/", auto_delete_objects=None if buckets_retention.upper() == "RETAIN" else True, id="raw-bucket", encryption=aws_s3.BucketEncryption.KMS_MANAGED if buckets_encryption_type.upper() == "KMS" else aws_s3.BucketEncryption.S3_MANAGED, block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, enforce_ssl=True, ) intermediate_bucket_name = f"{project_name}-{deployment_name}-intermediate-bucket-{hash}" unique_ab = (hashlib.sha256(module_name.encode("utf-8")).hexdigest())[: (60 - len(intermediate_bucket_name))] intermediate_bucket = aws_s3.Bucket( self, id="intermediate-bucket", removal_policy=aws_cdk.RemovalPolicy.RETAIN if buckets_retention.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, bucket_name=f"{intermediate_bucket_name}-{unique_ab}", versioned=True, server_access_logs_bucket=logs_bucket, server_access_logs_prefix="intermediate-bucket-logs/", auto_delete_objects=None if buckets_retention.upper() == "RETAIN" else True, encryption=aws_s3.BucketEncryption.KMS_MANAGED if buckets_encryption_type.upper() == "KMS" else aws_s3.BucketEncryption.S3_MANAGED, block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, enforce_ssl=True, ) curated_bucket_name = f"{project_name}-{deployment_name}-curated-bucket-{hash}" unique_ab = (hashlib.sha256(module_name.encode("utf-8")).hexdigest())[: (60 - len(curated_bucket_name))] curated_bucket = aws_s3.Bucket( self, id="curated-bucket", removal_policy=aws_cdk.RemovalPolicy.RETAIN if buckets_retention.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, bucket_name=f"{curated_bucket_name}-{unique_ab}", versioned=True, server_access_logs_bucket=logs_bucket, server_access_logs_prefix="curated-bucket-logs/", auto_delete_objects=None if buckets_retention.upper() == "RETAIN" else True, encryption=aws_s3.BucketEncryption.KMS_MANAGED if buckets_encryption_type.upper() == "KMS" else aws_s3.BucketEncryption.S3_MANAGED, block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, enforce_ssl=True, ) artifacts_bucket_name = f"{project_name}-{deployment_name}-artifacts-bucket-{hash}" unique_ab = (hashlib.sha256(module_name.encode("utf-8")).hexdigest())[: (60 - len(artifacts_bucket_name))] artifacts_bucket = aws_s3.Bucket( self, id="artifacts-bucket", bucket_name=f"{artifacts_bucket_name}-{unique_ab}", removal_policy=aws_cdk.RemovalPolicy.RETAIN if buckets_retention.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, auto_delete_objects=None if buckets_retention.upper() == "RETAIN" else True, encryption=aws_s3.BucketEncryption.KMS_MANAGED if buckets_encryption_type.upper() == "KMS" else aws_s3.BucketEncryption.S3_MANAGED, versioned=True, server_access_logs_bucket=logs_bucket, server_access_logs_prefix="artifacts-bucket-logs/", block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, enforce_ssl=True, ) # ReadOnly IAM Policy readonly_policy = aws_iam.ManagedPolicy( self, id="readonly_policy", managed_policy_name=f"{project_name}-{deployment_name}-{module_name}-{region}-{account}-readonly-access", statements=[ aws_iam.PolicyStatement( effect=aws_iam.Effect.ALLOW, actions=[ "kms:Decrypt", "kms:Encrypt", "kms:ReEncrypt*", "kms:DescribeKey", "kms:GenerateDataKey", ], resources=[f"arn:{partition}:kms::{account}:*"], ), aws_iam.PolicyStatement( effect=aws_iam.Effect.ALLOW, actions=[ "s3:GetObject", "s3:GetObjectAcl", "s3:ListBucket", ], resources=[ f"{raw_bucket.bucket_arn}/*", f"{raw_bucket.bucket_arn}", f"{intermediate_bucket.bucket_arn}/*", f"{intermediate_bucket.bucket_arn}", f"{curated_bucket.bucket_arn}/*", f"{curated_bucket.bucket_arn}", f"{artifacts_bucket.bucket_arn}/*", f"{artifacts_bucket.bucket_arn}", ], ), ], ) # FullAccess IAM Policy fullaccess_policy = aws_iam.ManagedPolicy( self, id="fullaccess_policy", managed_policy_name=f"{project_name}-{deployment_name}-{module_name}-{region}-{account}-full-access", statements=[ aws_iam.PolicyStatement( effect=aws_iam.Effect.ALLOW, actions=[ "kms:Decrypt", "kms:Encrypt", "kms:ReEncrypt*", "kms:DescribeKey", "kms:GenerateDataKey", ], resources=[f"arn:{partition}:kms::{account}:*"], ), aws_iam.PolicyStatement( effect=aws_iam.Effect.ALLOW, actions=[ "s3:GetObject", "s3:GetObjectAcl", "s3:ListBucket", ], resources=[ f"{raw_bucket.bucket_arn}/*", f"{raw_bucket.bucket_arn}", f"{intermediate_bucket.bucket_arn}/*", f"{intermediate_bucket.bucket_arn}", f"{curated_bucket.bucket_arn}/*", f"{curated_bucket.bucket_arn}", f"{artifacts_bucket.bucket_arn}/*", f"{artifacts_bucket.bucket_arn}", ], ), aws_iam.PolicyStatement( actions=["s3:PutObject", "s3:PutObjectAcl"], resources=[ f"{raw_bucket.bucket_arn}/*", f"{raw_bucket.bucket_arn}", f"{intermediate_bucket.bucket_arn}/*", f"{intermediate_bucket.bucket_arn}", f"{curated_bucket.bucket_arn}/*", f"{curated_bucket.bucket_arn}", f"{artifacts_bucket.bucket_arn}/*", f"{artifacts_bucket.bucket_arn}", ], ), ], ) self.raw_bucket = raw_bucket self.intermediate_bucket = intermediate_bucket self.curated_bucket = curated_bucket self.artifacts_bucket = artifacts_bucket self.logs_bucket = logs_bucket self.readonly_policy = readonly_policy self.fullaccess_policy = fullaccess_policy Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) suppressions = [ NagPackSuppression( **{ "id": "AwsSolutions-S1", "reason": "Logging has been disabled for demo purposes", } ), NagPackSuppression( **{ "id": "AwsSolutions-IAM5", "reason": "Resource access restriced for demo resources", } ), ] NagSuppressions.add_stack_suppressions(self, suppressions)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/optionals/datalake-buckets/app.py
modules/optionals/datalake-buckets/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import aws_cdk from aws_cdk import App, CfnOutput from stack import DataLakeBucketsStack # Project vars project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") hash = os.getenv("SEEDFARMER_HASH", "") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" # App Env vars buckets_encryption_type = os.getenv(_param("ENCRYPTION_TYPE"), "SSE") buckets_retention = os.getenv(_param("RETENTION_TYPE"), "DESTROY") artifact_logs_retention = os.getenv(_param("ARTIFACTS_LOG_RETENTION"), "1") if buckets_retention not in ["DESTROY", "RETAIN"]: raise ValueError("The only RETENTION_TYPE values accepted are 'DESTROY' and 'RETAIN' ") if buckets_encryption_type not in ["SSE", "KMS"]: raise ValueError("The only ENCRYPTION_TYPE values accepted are 'SSE' and 'KMS' ") def generate_description() -> str: soln_id = os.getenv(_param("SOLUTION_ID"), None) soln_name = os.getenv(_param("SOLUTION_NAME"), None) soln_version = os.getenv(_param("SOLUTION_VERSION"), None) desc = f"{project_name} - Datalake Buckets Module" if soln_id and soln_name and soln_version: desc = f"({soln_id}) {soln_name}. Version {soln_version}" elif soln_id and soln_name: desc = f"({soln_id}) {soln_name}" return desc app = App() stack = DataLakeBucketsStack( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, hash=hash, buckets_encryption_type=buckets_encryption_type, buckets_retention=buckets_retention, artifacts_log_retention=int(artifact_logs_retention), stack_description=generate_description(), env=aws_cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "ArtifactsBucketName": stack.artifacts_bucket.bucket_name, "LogsBucketName": stack.logs_bucket.bucket_name, "RawBucketName": stack.raw_bucket.bucket_name, "IntermediateBucketName": stack.intermediate_bucket.bucket_name, "CuratedBucketName": stack.curated_bucket.bucket_name, "ReadOnlyPolicyArn": stack.readonly_policy.managed_policy_arn, "FullAccessPolicyArn": stack.fullaccess_policy.managed_policy_arn, } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/optionals/datalake-buckets/tests/test_app.py
modules/optionals/datalake-buckets/tests/test_app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import pytest @pytest.fixture(scope="function") def stack_defaults(): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project" os.environ["SEEDFARMER_DEPLOYMENT_NAME"] = "test-deployment" os.environ["SEEDFARMER_MODULE_NAME"] = "test-module" os.environ["SEEDFARMER_HASH"] = "hash" os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" os.environ["SEEDFARMER_PARAMETER_ENCRYPTION_TYPE"] = "SSE" os.environ["SEEDFARMER_PARAMETER_RETENTION_TYPE"] = "DESTROY" # Unload the app import so that subsequent tests don't reuse if "app" in sys.modules: del sys.modules["app"] def test_app(stack_defaults): import app # noqa: F401 def test_buckets_encryption_type(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_ENCRYPTION_TYPE"] with pytest.raises(Exception): import app # noqa: F401 assert os.environ["SEEDFARMER_PARAMETER_ENCRYPTION_TYPE"] == "SSE" def test_invalid_buckets_encryption_type(stack_defaults): os.environ["SEEDFARMER_PARAMETER_ENCRYPTION_TYPE"] = "notvalid" with pytest.raises(Exception): import app # noqa: F401 def test_buckets_retention(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_RETENTION_TYPE"] with pytest.raises(Exception): import app # noqa: F401 assert os.environ["SEEDFARMER_PARAMETER_RETENTION_TYPE"] == "DESTROY" def test_invalid_retention_type(stack_defaults): os.environ["SEEDFARMER_PARAMETER_RETENTION_TYPE"] = "notvalid" with pytest.raises(Exception): import app # noqa: F401 def test_solution_description(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] = "v1.0.0" import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST. Version v1.0.0" def test_solution_description_no_version(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" del os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST"
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/optionals/datalake-buckets/tests/test_stack.py
modules/optionals/datalake-buckets/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" # Unload the app import so that subsequent tests don't reuse if "stack" in sys.modules: del sys.modules["stack"] def test_synthesize_stack(stack_defaults): import stack app = cdk.App() project_name = "test-project" dep_name = "test-deployment" mod_name = "test-module" metadata_storage_stack = stack.DataLakeBucketsStack( scope=app, id=f"{project_name}-{dep_name}-{mod_name}", project_name=project_name, deployment_name=dep_name, module_name=mod_name, hash="hash", buckets_encryption_type="SSE", buckets_retention="DESTROY", artifacts_log_retention=2, stack_description="Testing", env=cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) template = Template.from_stack(metadata_storage_stack) template.resource_count_is("AWS::S3::Bucket", 5)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/optionals/datalake-buckets/tests/__init__.py
modules/optionals/datalake-buckets/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/stack.py
modules/examples/example-spark-dags/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, cast import aws_cdk.aws_iam as iam import cdk_nag from aws_cdk import Aspects, Stack, Tags from cdk_nag import NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class DagIamRole(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, mwaa_exec_role: str, raw_bucket_name: str, **kwargs: Any, ) -> None: # Env vars self.project_name = project_name self.deployment_name = deployment_name self.module_name = module_name self.mwaa_exec_role = mwaa_exec_role super().__init__( scope, id, description="This stack deploys Example Spark DAGs resources", **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") # The below permissions is to deploy the `citibike` usecase declared in the below blogpost # https://aws.amazon.com/blogs/big-data/manage-and-process-your-big-data-workflows-with-amazon-mwaa-and-amazon-emr-on-amazon-eks/ policy_statements = [ iam.PolicyStatement( actions=["s3:ListBucket", "s3:GetObject*"], effect=iam.Effect.ALLOW, resources=["arn:aws:s3:::tripdata", "arn:aws:s3:::tripdata/*"], ), iam.PolicyStatement( actions=["s3:*"], effect=iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:s3:::{raw_bucket_name}", f"arn:{self.partition}:s3:::{raw_bucket_name}/*", ], ), iam.PolicyStatement( actions=[ "emr-containers:StartJobRun", "emr-containers:ListJobRuns", "emr-containers:DescribeJobRun", "emr-containers:CancelJobRun", ], effect=iam.Effect.ALLOW, resources=["*"], ), iam.PolicyStatement( actions=["kms:Decrypt", "kms:GenerateDataKey"], effect=iam.Effect.ALLOW, resources=[f"arn:{self.partition}:kms:{self.region}:{self.account}:key/*"], ), ] dag_document = iam.PolicyDocument(statements=policy_statements) r_name = f"{self.project_name}-{self.deployment_name}-{self.module_name}-dag-role" self.dag_role = iam.Role( self, f"dag-role-{self.deployment_name}-{self.module_name}", assumed_by=iam.ArnPrincipal(self.mwaa_exec_role), inline_policies={"DagPolicyDocument": dag_document}, role_name=r_name, path="/", ) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) NagSuppressions.add_stack_suppressions( self, [ { # type: ignore "id": "AwsSolutions-IAM5", "reason": "Resource access restriced describe only", }, ], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/app.py
modules/examples/example-spark-dags/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import aws_cdk from aws_cdk import App, CfnOutput from stack import DagIamRole project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") mwaa_exec_role = os.getenv("SEEDFARMER_PARAMETER_MWAA_EXEC_ROLE_ARN", "") raw_bucket_name = os.getenv("SEEDFARMER_PARAMETER_RAW_BUCKET_NAME", "") app = App() stack = DagIamRole( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, mwaa_exec_role=mwaa_exec_role, raw_bucket_name=raw_bucket_name, env=aws_cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string({"DagRoleArn": stack.dag_role.role_arn}), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/tests/test_stack.py
modules/examples/example-spark-dags/tests/test_stack.py
def test_placeholder() -> None: return None
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/tests/__init__.py
modules/examples/example-spark-dags/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/example_spark_dags/citibike-spark-all.py
modules/examples/example-spark-dags/example_spark_dags/citibike-spark-all.py
""" Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. """ import sys from pyspark.sql import SparkSession from pyspark.sql.types import ( DateType, DoubleType, IntegerType, StringType, StructField, StructType, ) bucket = sys.argv[1] spark = SparkSession.builder.appName("citibike").getOrCreate() schema = StructType( [ StructField("tripduration", IntegerType(), True), StructField("starttime", DateType(), True), StructField("stoptime", DateType(), True), StructField("start_station_id", IntegerType(), True), StructField("start_station_name", StringType(), True), StructField("start_station_lat", DoubleType(), True), StructField("start_station_lon", DoubleType(), True), StructField("end_station_id", IntegerType(), True), StructField("end_station_name", StringType(), True), StructField("end_station_lat", DoubleType(), True), StructField("end_station_lon", DoubleType(), True), StructField("bike_id", IntegerType(), True), StructField("usertype", StringType(), True), StructField("birthyear", IntegerType(), True), StructField("gender", IntegerType(), True), ] ) df = spark.read.format("csv").option("header", "true").schema(schema).load("s3://" + bucket + "/citibike/csv/") df.write.parquet("s3://" + bucket + "/citibike/parquet/", mode="overwrite") ####################################### df = spark.read.format("parquet").load("s3://" + bucket + "/citibike/parquet/") df.createOrReplaceTempView("citibike") newdf = spark.sql("select *, extract(year from starttime) as yr,extract( month from starttime) as mo from citibike") newdf.createOrReplaceTempView("newcitibike") output = spark.sql("select count(*) as sum, mo from newcitibike group by mo order by sum desc") output.coalesce(1).write.format("csv").option("header", "true").save( "s3://" + bucket + "/citibike/results/ridership/", mode="overwrite" ) output = spark.sql( "select start_station_name,count(start_station_name) as counts, mo from newcitibike group by start_station_name,mo order by counts desc" ) output.coalesce(1).write.format("csv").option("header", "true").save( "s3://" + bucket + "/citibike/results/popular_start_stations/", mode="overwrite" ) output = spark.sql( "select end_station_name,count(end_station_name) as counts, mo from newcitibike group by end_station_name,mo order by counts desc" ) output.coalesce(1).write.format("csv").option("header", "true").save( "s3://" + bucket + "/citibike/results/popular_end_stations/", mode="overwrite" ) output = spark.sql( "select count(birthyear) as trips,mo as month,2021-birthyear as age from newcitibike group by age,month order by trips desc" ) output.coalesce(1).write.format("csv").option("header", "true").save( "s3://" + bucket + "/citibike/results/trips_by_age/", mode="overwrite" )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/example_spark_dags/citibike_all_dag.py
modules/examples/example-spark-dags/example_spark_dags/citibike_all_dag.py
""" Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. """ import json import random import string import zipfile from datetime import datetime, timedelta from io import BytesIO import boto3 from airflow import DAG, settings from airflow.contrib.hooks.aws_hook import AwsHook from airflow.exceptions import AirflowException from airflow.models import Connection from airflow.operators.dummy_operator import DummyOperator from airflow.operators.python_operator import PythonOperator from airflow.providers.amazon.aws.operators.emr_containers import EMRContainerOperator from airflow.providers.amazon.aws.sensors.emr_containers import EMRContainerSensor from boto3.session import Session from example_spark_dags import emr_eks_dag_config afbucket = f"{emr_eks_dag_config.DAG_BUCKET}/dags/example_spark_dags/" # ADDF MWAA Dags bucket emr_virtual_cluster_id = emr_eks_dag_config.VIRTUAL_CLUSTER_ID emr_execution_role_arn = emr_eks_dag_config.EMR_JOB_EXECUTION_ROLE YR = "2020" bucket = emr_eks_dag_config.RAW_BUCKET # ADDF RAW bucket SRC_BUCKET = "tripdata" SRC_KEY = "-citibike-tripdata.csv.zip" DEST_BUCKET = emr_eks_dag_config.RAW_BUCKET DEST_KEY = "-citibike-tripdata.csv" now = datetime.now() def try_create_aws_conn(**kwargs): conn_id = "aws_emr_on_eks" try: AwsHook.get_connection(conn_id) except AirflowException: extra = json.dumps({"role_arn": emr_eks_dag_config.DAG_ROLE}, indent=2) conn = Connection(conn_id=conn_id, conn_type="aws", host="", schema="", login="", extra=extra) try: session = settings.Session() session.add(conn) session.commit() finally: session.close() def get_assumerole_creds(): sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=emr_eks_dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session def find_max_month(): session = get_assumerole_creds() conn = session.client("s3") mo = 0 for key in conn.list_objects(Bucket=SRC_BUCKET)["Contents"]: if "JC" not in key["Key"] and YR in key["Key"]: mo = mo + 1 print("returning max month {}".format(mo)) return mo def copy_and_unzip_s3(**context): session = get_assumerole_creds() s3_resource = session.resource("s3") zip_obj = s3_resource.Object(bucket_name=context["bucket"], key=context["key"]) buffer = BytesIO(zip_obj.get()["Body"].read()) z = zipfile.ZipFile(buffer) print("downloaded zip {}, zipObj {}".format(z, zipfile)) for filename in z.namelist(): if filename.startswith("__"): continue file_info = z.getinfo(filename) print("interating over zip {}, zipObj {}".format(filename, file_info)) try: response = s3_resource.meta.client.upload_fileobj( z.open(filename), Bucket=context["destbucket"], Key=context["destkey"] ) print("uploaded to s3 {}".format(filename)) except Exception as e: print(e) def list_bucket(**context): session = get_assumerole_creds() conn = session.client("s3") for key in conn.list_objects(Bucket=context["destbucket"])["Contents"]: if "csv.gz" in key["Key"]: print(key["Key"]) DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], "email_on_failure": False, "email_on_retry": False, "retries": 5, "retry_delay": timedelta(minutes=1), } # [START howto_operator_emr_containers_start_job_run] JOB_DRIVER = { "sparkSubmitJobDriver": { "entryPoint": "s3://" + afbucket + "citibike-spark-all.py", "entryPointArguments": [bucket], "sparkSubmitParameters": "--conf spark.executor.instances=3 --conf " "spark.executor.memory=4G --conf spark.driver.memory=2G --conf spark.executor.cores=2 " "--conf spark.sql.shuffle.partitions=60 --conf spark.dynamicAllocation.enabled=false", } } CONFIGURATION_OVERRIDES = { "monitoringConfiguration": { "cloudWatchMonitoringConfiguration": { "logGroupName": "/emr-containers/jobs", "logStreamNamePrefix": "addf", }, "persistentAppUI": "ENABLED", "s3MonitoringConfiguration": {"logUri": "s3://" + afbucket + "/joblogs"}, } } with DAG( dag_id="Citibike_Ridership_Analytics", default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=datetime(now.year, now.month, now.day, now.hour), schedule_interval=None, tags=["S3", "Citibike", "EMR on EKS", "Spark"], ) as dag: start = DummyOperator(task_id="start", dag=dag) listBucket = PythonOperator( task_id="list_transformed_files", python_callable=list_bucket, op_kwargs={"destbucket": DEST_BUCKET}, dag=dag, ) create_aws_conn = PythonOperator( task_id="try_create_aws_conn", python_callable=try_create_aws_conn, dag=dag, ) for i in range(1, find_max_month() + 1): NEW_SRC_KEY = YR + str(i).zfill(2) + SRC_KEY NEW_DEST_KEY = "citibike/csv/" + YR + str(i).zfill(2) + DEST_KEY copyAndTransformS3File = PythonOperator( task_id="copy_and_unzip_s3_" + str(i).zfill(2), python_callable=copy_and_unzip_s3, op_kwargs={ "bucket": SRC_BUCKET, "key": NEW_SRC_KEY, "destbucket": DEST_BUCKET, "destkey": NEW_DEST_KEY, }, dag=dag, ) start >> copyAndTransformS3File >> listBucket start_job_run = EMRContainerOperator( task_id=f"start_citibike_ridership_analytics-{i}", name="citibike_analytics_run", virtual_cluster_id=emr_virtual_cluster_id, client_request_token="".join(random.choice(string.digits) for _ in range(10)), execution_role_arn=emr_execution_role_arn, release_label="emr-6.2.0-latest", job_driver=JOB_DRIVER, configuration_overrides=CONFIGURATION_OVERRIDES, aws_conn_id="aws_emr_on_eks", ) job_sensor = EMRContainerSensor( task_id=f"check_job_status-{i}", job_id="{{ task_instance.xcom_pull(task_ids='start_citibike_ridership_analytics', key='return_value') }}", virtual_cluster_id="{{ task_instance.xcom_pull(task_ids='start_citibike_ridership_analytics', key='virtual_cluster_id') }}", aws_conn_id="aws_emr_on_eks", ) create_aws_conn >> listBucket >> start_job_run >> job_sensor
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/example_spark_dags/__init__.py
modules/examples/example-spark-dags/example_spark_dags/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-spark-dags/example_spark_dags/emr_eks_dag_config.py
modules/examples/example-spark-dags/example_spark_dags/emr_eks_dag_config.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 # This file is populated with configurations information when the Module is deployed # Configuration parameters are exported as module level constants # # Example: # SOME_PARAMETER = 'some value'
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/docker-build/service/app/app.py
modules/examples/docker-build/service/app/app.py
from flask import Flask app = Flask(__name__) @app.route("/") def hello_world(): return "Hello, Docker!"
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/eb-sf-batch/stack.py
modules/examples/eb-sf-batch/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, cast import aws_cdk.aws_batch as batch import aws_cdk.aws_iam as iam from aws_cdk import Duration, NestedStack, Size, Stack, Tags from aws_cdk import aws_ecr as ecr from aws_cdk import aws_ecs as ecs from aws_cdk import aws_events as events from aws_cdk import aws_stepfunctions as stepfunctions from aws_cdk import aws_stepfunctions_tasks as step_functions_tasks from aws_solutions_constructs.aws_eventbridge_stepfunctions import ( EventbridgeToStepfunctions, ) # from cdk_nag import NagSuppressions from constructs import Construct, IConstruct # import cdk_nag _logger: logging.Logger = logging.getLogger(__name__) class EventDrivenBatch(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, fargate_job_queue_arn: str, ecr_repo_name: str, vcpus: int, memory_limit_mib: int, **kwargs: Any, ) -> None: # Env vars self.deployment_name = deployment_name self.module_name = module_name super().__init__( scope, id, description="This stack deploys Cron Based Eventbridge which triggers Stepfunctions further triggering AWS Batch", # noqa: E501 **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") dep_mod = f"{project_name}-{deployment_name}-{module_name}" # Batch Resources role = iam.Role( self, f"{dep_mod}-AWSBatchRole", assumed_by=iam.ServicePrincipal(service="ecs-tasks.amazonaws.com"), managed_policies=[ iam.ManagedPolicy.from_managed_policy_arn( self, "MetricsAmazonEC2ContainerRegistryReadOnly", f"arn:{self.partition}:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly", ), iam.ManagedPolicy.from_managed_policy_arn( self, "MetricsAmazonEC2ContainerServiceforEC2Role", f"arn:{self.partition}:iam::aws:policy/service-role/AmazonEC2ContainerServiceforEC2Role", ), ], inline_policies={ "S3Read": iam.PolicyDocument( statements=[ iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=["s3:GetObject"], resources=["*"], ) ] ) }, ) repository = ecr.Repository.from_repository_name(self, id=id, repository_name=f"{dep_mod}-{ecr_repo_name}") img = ecs.EcrImage.from_ecr_repository(repository=repository, tag="latest") definition = batch.EcsJobDefinition( self, f"{dep_mod}-JobDefinition", job_definition_name=f"{project_name}-{deployment_name}-Job-Definition", retry_attempts=1, container=batch.EcsFargateContainerDefinition( self, f"{dep_mod} Container Definition", environment={"AWS_REGION": NestedStack.of(self).region}, cpu=int(vcpus), memory=Size.mebibytes(int(memory_limit_mib)), execution_role=role, job_role=role, image=img, # command=["echo", "I ran fine"], ), ) # Step functions Definition submit_metrics_job = step_functions_tasks.BatchSubmitJob( self, f"{dep_mod}-Batchjob", job_name=f"{project_name}-{deployment_name}-Job", job_queue_arn=fargate_job_queue_arn, job_definition_arn=definition.job_definition_arn, ) wait_job = stepfunctions.Wait( self, "Wait 30 Seconds", time=stepfunctions.WaitTime.duration(Duration.seconds(30)), ) # fail_job = stepfunctions.Fail(self, "Fail", cause="AWS Batch Job Failed", error="DescribeJob returned FAILED") succeed_job = stepfunctions.Succeed(self, "Succeeded", comment="AWS Batch Job succeeded") # Create Chain definition = submit_metrics_job.next(wait_job).next(succeed_job) # type: ignore self.eventbridge_sfn = EventbridgeToStepfunctions( self, f"{project_name}-{deployment_name}-eb-sf-batch", state_machine_props=stepfunctions.StateMachineProps(definition=definition), # type: ignore event_rule_props=events.RuleProps(schedule=events.Schedule.rate(Duration.minutes(1))), )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/eb-sf-batch/app.py
modules/examples/eb-sf-batch/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from aws_cdk import App, CfnOutput, Environment from stack import EventDrivenBatch project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" vcpus = os.getenv(_param("VCPUS")) # required memory_limit_mib = os.getenv(_param("MEMORY_LIMIT_MIB")) # required fargate_job_queue_arn = os.getenv(_param("FARGATE_JOB_QUEUE_ARN")) # required ecr_repo_name = os.getenv(_param("ECR_REPO_NAME")) # required if not fargate_job_queue_arn: raise ValueError("Batch Queue Configuration is missing.") if not ecr_repo_name: raise ValueError("ECR Repository Name is missing.") app = App() stack = EventDrivenBatch( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, fargate_job_queue_arn=fargate_job_queue_arn, ecr_repo_name=ecr_repo_name, vcpus=vcpus, # type: ignore memory_limit_mib=memory_limit_mib, # type: ignore env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "StateMachine": stack.eventbridge_sfn.state_machine.state_machine_arn, } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/eb-sf-batch/tests/test_stack.py
modules/examples/eb-sf-batch/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project" os.environ["SEEDFARMER_DEPLOYMENT_NAME"] = "test-deployment" os.environ["SEEDFARMER_MODULE_NAME"] = "test-module" os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" if "stack" in sys.modules: del sys.modules["stack"] def test_synthesize_stack(stack_defaults): import stack app = cdk.App() project_name = "test-project" dep_name = "test-deployment" mod_name = "test-module" event_driven_batch = stack.EventDrivenBatch( scope=app, id=f"{project_name}-{dep_name}-{mod_name}", project_name=project_name, deployment_name=dep_name, module_name=mod_name, fargate_job_queue_arn="arn:foobar", ecr_repo_name="myrepo", vcpus=2, memory_limit_mib=2048, env=cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) template = Template.from_stack(event_driven_batch) template.resource_count_is("AWS::Batch::JobDefinition", 1)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/eb-sf-batch/tests/__init__.py
modules/examples/eb-sf-batch/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/eb-sf-batch/images/src.py
modules/examples/eb-sf-batch/images/src.py
import pandas as pd # initialize list of lists data = [["Cristiano Ronlado", 1], ["Juan Mata", 2], ["Bruno Fernandez", 3]] # Create the pandas DataFrame df = pd.DataFrame(data, columns=["Name", "Jersey Number"]) print("-" * 50) print("StepFunctions - eventbridge - AWS Batch workflow tested") print("-" * 50) print(df) print("-" * 50)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-tf-prereqs/stack.py
modules/examples/example-tf-prereqs/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import logging from typing import Any, cast import aws_cdk import cdk_nag from aws_cdk import Aspects, Stack, Tags, aws_dynamodb, aws_s3 from cdk_nag import NagPackSuppression, NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class TfPreReqs(Stack): # type: ignore def __init__( self, scope: Construct, id: str, project_name: str, deployment_name: str, module_name: str, hash: str, tf_s3_backend_encryption_type: str, tf_s3_backend_retention_type: str, tf_ddb_retention_type: str, **kwargs: Any, ) -> None: super().__init__( scope, id, description="This stack deploys Storage resources", **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") # S3 bucket for storing the remote state of Terraform self.tf_state_s3bucket = aws_s3.Bucket( self, removal_policy=aws_cdk.RemovalPolicy.RETAIN if tf_s3_backend_retention_type.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, bucket_name=f"{project_name}-{deployment_name}-tfstate-bucket-{hash}", id="tf-state-bucket", encryption=aws_s3.BucketEncryption.KMS_MANAGED if tf_s3_backend_encryption_type.upper() == "KMS" else aws_s3.BucketEncryption.S3_MANAGED, block_public_access=aws_s3.BlockPublicAccess.BLOCK_ALL, enforce_ssl=True, versioned=True, ) # DDB Table for storing the LockIDs of Terraform part_key = "LockID" self.tf_ddb_lock_table = aws_dynamodb.Table( self, "tf_ddb_lock_table", table_name=f"{project_name}-{deployment_name}-tf-ddb-lock-table", partition_key=aws_dynamodb.Attribute(name=part_key, type=aws_dynamodb.AttributeType.STRING), billing_mode=aws_dynamodb.BillingMode.PAY_PER_REQUEST, removal_policy=aws_cdk.RemovalPolicy.RETAIN if tf_ddb_retention_type.upper() == "RETAIN" else aws_cdk.RemovalPolicy.DESTROY, ) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) suppressions = [ NagPackSuppression( **{ "id": "AwsSolutions-S1", "reason": "Logging has been disabled for demo purposes", } ), NagPackSuppression( **{ "id": "AwsSolutions-IAM5", "reason": "Resource access restriced to ADDF resources", } ), ] NagSuppressions.add_stack_suppressions(self, suppressions)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-tf-prereqs/app.py
modules/examples/example-tf-prereqs/app.py
import os import aws_cdk from aws_cdk import App, CfnOutput from stack import TfPreReqs project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") hash = os.getenv("SEEDFARMER_HASH", "") tf_s3_backend_encryption_type = os.getenv("SEEDFARMER_PARAMETER_S3_ENCRYPTION_TYPE", "SSE") tf_s3_backend_retention_type = os.getenv("SEEDFARMER_PARAMETER_S3_RETENTION_TYPE", "DESTROY") tf_ddb_retention_type = os.getenv("SEEDFARMER_PARAMETER_DDB_RETENTION_TYPE", "DESTROY") app = App() stack = TfPreReqs( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, hash=hash, tf_s3_backend_encryption_type=tf_s3_backend_encryption_type, tf_s3_backend_retention_type=tf_s3_backend_retention_type, tf_ddb_retention_type=tf_ddb_retention_type, env=aws_cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "TfStateBucketName": stack.tf_state_s3bucket.bucket_name, "TfLockTable": stack.tf_ddb_lock_table.table_name, } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-tf-prereqs/tests/test_stack.py
modules/examples/example-tf-prereqs/tests/test_stack.py
def test_placeholder() -> None: return None
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-tf-prereqs/tests/__init__.py
modules/examples/example-tf-prereqs/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/stack.py
modules/examples/example-dags/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, Optional, cast import aws_cdk.aws_iam as aws_iam import cdk_nag from aws_cdk import Aspects, Stack, Tags from cdk_nag import NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class DagIamRole(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, mwaa_exec_role: str, bucket_policy_arn: Optional[str] = None, permission_boundary_arn: Optional[str] = None, **kwargs: Any, ) -> None: # Env vars self.project_name = project_name self.deployment_name = deployment_name self.module_name = module_name self.mwaa_exec_role = mwaa_exec_role super().__init__( scope, id, description="This stack deploys Example DAGs resources", **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") # Create Dag IAM Role and policy dag_statement = aws_iam.PolicyDocument( statements=[ aws_iam.PolicyStatement( actions=["ec2:Describe*"], effect=aws_iam.Effect.ALLOW, resources=["*"], ) ] ) managed_policies = ( [aws_iam.ManagedPolicy.from_managed_policy_arn(self, "bucket-policy", bucket_policy_arn)] if bucket_policy_arn else [] ) # Role with Permission Boundary r_name = f"{self.project_name}-{self.deployment_name}-{self.module_name}-dag-role" self.dag_role = aws_iam.Role( self, f"dag-role-{self.deployment_name}-{self.module_name}", assumed_by=aws_iam.ArnPrincipal(self.mwaa_exec_role), inline_policies={"DagPolicyDocument": dag_statement}, managed_policies=managed_policies, permissions_boundary=aws_iam.ManagedPolicy.from_managed_policy_arn( self, f"perm-boundary-{self.deployment_name}-{self.module_name}", permission_boundary_arn, ) if permission_boundary_arn else None, role_name=r_name, path="/", ) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) NagSuppressions.add_stack_suppressions( self, [ { # type: ignore "id": "AwsSolutions-IAM5", "reason": "Resource access restriced describe only", }, ], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/app.py
modules/examples/example-dags/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import aws_cdk from aws_cdk import App, CfnOutput from stack import DagIamRole project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") mwaa_exec_role = os.getenv("SEEDFARMER_PARAMETER_MWAA_EXEC_ROLE_ARN", "") bucket_policy_arn = os.getenv("SEEDFARMER_PARAMETER_BUCKET_POLICY_ARN") permission_boundary_arn = os.getenv("SEEDFARMER_PERMISSION_BOUNDARY_ARN") app = App() stack = DagIamRole( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, mwaa_exec_role=mwaa_exec_role, bucket_policy_arn=bucket_policy_arn, permission_boundary_arn=permission_boundary_arn, env=aws_cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string({"DagRoleArn": stack.dag_role.role_arn}), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/tests/test_stack.py
modules/examples/example-dags/tests/test_stack.py
def test_placeholder() -> None: return None
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/tests/__init__.py
modules/examples/example-dags/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/example_dags/dag_config.py
modules/examples/example-dags/example_dags/dag_config.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/example_dags/sampledag.py
modules/examples/example-dags/example_dags/sampledag.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from datetime import timedelta import boto3 from airflow import DAG from airflow.operators.python_operator import PythonOperator from airflow.utils.dates import days_ago from boto3.session import Session import example_dags.dag_config as cf DAG_ID = os.path.basename(__file__).replace(".py", "") DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } def triggerDagFn(**kwargs): sts_client = boto3.client("sts") # Call the assume_role method of the STSConnection object and pass the role # ARN and a role session name response = sts_client.assume_role( RoleArn=cf.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) new_sts_client = session.client("sts") print(f"The new client is : {new_sts_client.get_caller_identity()}") ec2_client = session.client("ec2") response = ec2_client.describe_instances() print(f"response is: {response}") return True with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=days_ago(1), schedule_interval="@once", ) as dag: triggerDag = PythonOperator(task_id="triggerDag", python_callable=triggerDagFn, provide_context=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/examples/example-dags/example_dags/__init__.py
modules/examples/example-dags/example_dags/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/core/metadata-storage/stack.py
modules/core/metadata-storage/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, cast import cdk_nag from aws_cdk import Aspects, RemovalPolicy, Stack, Tags from aws_cdk import aws_dynamodb as dynamo from aws_cdk import aws_glue_alpha as glue_alpha from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class MetadataStorageStack(Stack): def __init__( self, scope: Construct, id: str, project_name: str, deployment_name: str, module_name: str, scene_table_suffix: str, bagfile_table_suffix: str, glue_db_suffix: str, stack_description: str, **kwargs: Any, ) -> None: dep_mod = f"{project_name}-{deployment_name}-{module_name}" # used to tag AWS resources. Tag Value length cant exceed 256 characters full_dep_mod = dep_mod[:256] if len(dep_mod) > 256 else dep_mod super().__init__(scope, id, description=stack_description, **kwargs) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=full_dep_mod) rosbag_bagfile_p_key = "bag_file_prefix" rosbag_bagfile_table = dynamo.Table( self, "dynamobagfiletable", table_name=f"{full_dep_mod}-{bagfile_table_suffix}", partition_key=dynamo.Attribute(name=rosbag_bagfile_p_key, type=dynamo.AttributeType.STRING), billing_mode=dynamo.BillingMode.PAY_PER_REQUEST, removal_policy=RemovalPolicy.DESTROY, point_in_time_recovery=True, ) rosbag_scene_p_key = "bag_file" rosbag_scene_sort_key = "scene_id" rosbag_scene_table = dynamo.Table( self, "dynamotablescenes", table_name=f"{full_dep_mod}-{scene_table_suffix}", partition_key=dynamo.Attribute(name=rosbag_scene_p_key, type=dynamo.AttributeType.STRING), sort_key=dynamo.Attribute(name=rosbag_scene_sort_key, type=dynamo.AttributeType.STRING), billing_mode=dynamo.BillingMode.PAY_PER_REQUEST, removal_policy=RemovalPolicy.DESTROY, point_in_time_recovery=True, stream=dynamo.StreamViewType.NEW_AND_OLD_IMAGES, ) glue_db = glue_alpha.Database( self, "glue_db", database_name=f"{full_dep_mod}-{glue_db_suffix}", ) self.rosbag_bagfile_table = rosbag_bagfile_table self.rosbag_bagfile_partition_key = rosbag_bagfile_p_key self.rosbag_scene_table = rosbag_scene_table self.rosbag_scene_table_stream_arn = rosbag_scene_table.table_stream_arn self.rosbag_scene_p_key = rosbag_scene_p_key self.rosbag_scene_sort_key = rosbag_scene_sort_key self.glue_db = glue_db Aspects.of(self).add(cdk_nag.AwsSolutionsChecks(verbose=True))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/core/metadata-storage/app.py
modules/core/metadata-storage/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import aws_cdk from aws_cdk import App, CfnOutput from stack import MetadataStorageStack # Project vars project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" # App Env vars scene_suffix = os.getenv(_param("ROSBAG_SCENE_TABLE_SUFFIX")) if not scene_suffix: raise ValueError("ROSBAG_SCENE_TABLE_SUFFIX is not populated ") glue_db_suffix = os.getenv(_param("GLUE_DB_SUFFIX")) if not glue_db_suffix: raise ValueError("GLUE_DB_SUFFIX is not populated") bagfile_suffix = os.getenv(_param("ROSBAG_BAGFILE_TABLE_SUFFIX")) if not bagfile_suffix: raise ValueError("ROSBAG_BAGFILE_TABLE_SUFFIX is not populated") def generate_description() -> str: soln_id = os.getenv(_param("SOLUTION_ID"), None) soln_name = os.getenv(_param("SOLUTION_NAME"), None) soln_version = os.getenv(_param("SOLUTION_VERSION"), None) desc = f"{project_name} - Metadata Storage Module" if soln_id and soln_name and soln_version: desc = f"({soln_id}) {soln_name}. Version {soln_version}" elif soln_id and soln_name: desc = f"({soln_id}) {soln_name}" return desc app = App() stack = MetadataStorageStack( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", env=aws_cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), project_name=project_name, deployment_name=deployment_name, module_name=module_name, scene_table_suffix=scene_suffix, bagfile_table_suffix=bagfile_suffix, glue_db_suffix=glue_db_suffix, stack_description=generate_description(), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "RosbagBagFileTable": stack.rosbag_bagfile_table.table_name, "RosbagBagFilePartitionKey": stack.rosbag_bagfile_partition_key, "RosbagSceneMetadataTable": stack.rosbag_scene_table.table_name, "RosbagSceneMetadataStreamArn": stack.rosbag_scene_table_stream_arn, "RosbagSceneMetadataPartitionKey": stack.rosbag_scene_p_key, "RosbagSceneMetadataSortKey": stack.rosbag_scene_sort_key, "GlueDBName": stack.glue_db.database_name, } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/core/metadata-storage/tests/test_app.py
modules/core/metadata-storage/tests/test_app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import pytest @pytest.fixture(scope="function") def stack_defaults(): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project" os.environ["SEEDFARMER_DEPLOYMENT_NAME"] = "test-deployment" os.environ["SEEDFARMER_MODULE_NAME"] = "test-module" os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" os.environ["SEEDFARMER_PARAMETER_ROSBAG_SCENE_TABLE_SUFFIX"] = "scene-suffix" os.environ["SEEDFARMER_PARAMETER_GLUE_DB_SUFFIX"] = "glue-db-suffix" os.environ["SEEDFARMER_PARAMETER_ROSBAG_BAGFILE_TABLE_SUFFIX"] = "ros-table-suffix" # Unload the app import so that subsequent tests don't reuse if "app" in sys.modules: del sys.modules["app"] def test_app(stack_defaults): import app # noqa: F401 def test_scene_suffix(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_ROSBAG_SCENE_TABLE_SUFFIX"] with pytest.raises(Exception): import app # noqa: F401 assert os.environ["SEEDFARMER_PARAMETER_ROSBAG_SCENE_TABLE_SUFFIX"] == "scene-suffix" def test_glue_db_suffix(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_GLUE_DB_SUFFIX"] with pytest.raises(Exception): import app # noqa: F401 assert os.environ["SEEDFARMER_PARAMETER_GLUE_DB_SUFFIX"] == "glue-db-suffix" def test_bagfile_suffix(stack_defaults): del os.environ["SEEDFARMER_PARAMETER_ROSBAG_BAGFILE_TABLE_SUFFIX"] with pytest.raises(Exception): import app # noqa: F401 assert os.environ["SEEDFARMER_PARAMETER_ROSBAG_BAGFILE_TABLE_SUFFIX"] == "ros-table-suffix" def test_solution_description(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] = "v1.0.0" import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST. Version v1.0.0" def test_solution_description_no_version(stack_defaults): os.environ["SEEDFARMER_PARAMETER_SOLUTION_ID"] = "SO123456" os.environ["SEEDFARMER_PARAMETER_SOLUTION_NAME"] = "MY GREAT TEST" del os.environ["SEEDFARMER_PARAMETER_SOLUTION_VERSION"] import app ver = app.generate_description() assert ver == "(SO123456) MY GREAT TEST"
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/core/metadata-storage/tests/test_stack.py
modules/core/metadata-storage/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["CDK_DEFAULT_ACCOUNT"] = "111111111111" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" # Unload the app import so that subsequent tests don't reuse if "stack" in sys.modules: del sys.modules["stack"] def test_synthesize_stack(stack_defaults): import stack app = cdk.App() project_name = "test-project" deployment_name = "test-deployment" module_name = "test-module" metadata_storage_stack = stack.MetadataStorageStack( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, scene_table_suffix="scene-suffix", bagfile_table_suffix="glue-db-suffix", glue_db_suffix="ros-table-suffix", stack_description="Testing", env=cdk.Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) template = Template.from_stack(metadata_storage_stack) template.resource_count_is("AWS::DynamoDB::Table", 2) template.resource_count_is("AWS::Glue::Database", 1)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/core/metadata-storage/tests/__init__.py
modules/core/metadata-storage/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/eureka/stack.py
modules/simulations/eureka/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from string import Template from typing import Any, cast import yaml from aws_cdk import Duration, Environment, Stack, Tags from aws_cdk import aws_ecr as ecr from aws_cdk import aws_eks as eks from aws_cdk import aws_iam as iam from aws_cdk import aws_sqs as sqs from aws_cdk.aws_ecr_assets import DockerImageAsset from aws_cdk.lambda_layer_kubectl_v29 import KubectlV29Layer from cdk_ecr_deployment import DockerImageName, ECRDeployment from cdk_nag import NagPackSuppression, NagSuppressions from constructs import Construct, IConstruct project_dir = os.path.dirname(os.path.abspath(__file__)) ROLE_NAME = "addf-eureka-simulation-role" APPLICATION_IMAGE_NAME = "ubuntu-ros2" class EurekaStack(Stack): def __init__( self, scope: Construct, id: str, project_name: str, deployment_name: str, module_name: str, stack_description: str, eks_cluster_name: str, eks_cluster_admin_role_arn: str, eks_oidc_arn: str, eks_cluster_open_id_connect_issuer: str, simulation_data_bucket_name: str, sqs_name: str, fsx_volume_handle: str, fsx_mount_point: str, application_ecr_name: str, env: Environment, **kwargs: Any, ) -> None: super().__init__(scope, id, description=stack_description, **kwargs) self.project_name = project_name self.deployment_name = deployment_name self.module_name = module_name dep_mod = f"{self.project_name}-{self.deployment_name}-{self.module_name}" dep_mod = dep_mod[:64] Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=dep_mod) provider = eks.OpenIdConnectProvider.from_open_id_connect_provider_arn( self, f"{dep_mod}-provider", eks_oidc_arn ) eks_cluster = eks.Cluster.from_cluster_attributes( self, f"{dep_mod}-eks-cluster", cluster_name=eks_cluster_name, kubectl_role_arn=eks_cluster_admin_role_arn, open_id_connect_provider=provider, kubectl_layer=KubectlV29Layer(self, "Kubectlv29Layer"), ) manifest = self.get_fsx_static_provisioning_manifest(fsx_volume_handle, fsx_mount_point, env) manifest_file = list(yaml.load_all(manifest, Loader=yaml.FullLoader)) loop_iteration = 0 for value in manifest_file: loop_iteration = loop_iteration + 1 manifest_id = "Eureka" + str(loop_iteration) eks_cluster.add_manifest(manifest_id, value) self.iam_role_arn = self.create_simulation_role( eks_cluster_open_id_connect_issuer, eks_oidc_arn, simulation_data_bucket_name, ) self.sqs_url = self.create_sqs(sqs_name, env) self.application_image_uri = self.build_application_image(application_ecr_name) NagSuppressions.add_stack_suppressions( self, apply_to_nested_stacks=True, suppressions=[ NagPackSuppression( **{ "id": "AwsSolutions-IAM4", "reason": "Managed Policies are for service account roles only", } ), NagPackSuppression( **{ "id": "AwsSolutions-IAM5", "reason": "Resource access restriced to resources", } ), ], ) def create_simulation_role( self, eks_cluster_open_id_connect_issuer: str, eks_oidc_arn: str, simulation_data_bucket_name: str, ) -> str: role = iam.Role( self, ROLE_NAME, assumed_by=iam.FederatedPrincipal( eks_oidc_arn, {"StringLike": {f"{eks_cluster_open_id_connect_issuer}:sub": "system:serviceaccount:*"}}, "sts:AssumeRoleWithWebIdentity", ), ) role.add_to_principal_policy( iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=[ "s3:PutObject", "s3:GetObject", "s3:AbortMultipartUpload", "s3:ListBucket", "s3:GetObjectVersion", "s3:ListMultipartUploadParts", ], resources=[ f"arn:{self.partition}:s3:::${simulation_data_bucket_name}/*", f"arn:{self.partition}:s3:::${simulation_data_bucket_name}", ], ) ) role.add_to_principal_policy( iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=[ "sqs:*", "fsx:CreateFileSystem", "fsx:DeleteFileSystem", "fsx:DescribeFileSystems", "fsx:TagResource", ], resources=["*"], ) ) role.add_to_principal_policy( iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=[ "bedrock:ListFoundationModels", "bedrock:InvokeModelWithResponseStream", "bedrock:InvokeModel", "bedrock:InvokeAgent", ], resources=["*"], ) ) role.add_to_principal_policy( iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=[ "iam:CreateServiceLinkedRole", "iam:AttachRolePolicy", "iam:PutRolePolicy", ], resources=[ f"arn:{self.partition}:iam::*:role/aws-service-role/s3.data-source.lustre.fsx.amazonaws.com/*" ], ) ) fsx_policy = iam.PolicyStatement( effect=iam.Effect.ALLOW, actions=[ "iam:CreateServiceLinkedRole", ], resources=["*"], ) fsx_policy.add_conditions({"StringLike": {"iam:AWSServiceName": ["fsx.amazonaws.com"]}}) role.add_to_principal_policy(fsx_policy) return role.role_arn def build_application_image(self, application_ecr_name: str) -> str: local_image = DockerImageAsset( self, "ImageExtractionDockerImage", directory=os.path.join(os.path.dirname(os.path.abspath(__file__)), "src"), ) repo = ecr.Repository.from_repository_name( self, id=f"ecr-{application_ecr_name}", repository_name=application_ecr_name ) image_uri = f"{repo.repository_uri}:{APPLICATION_IMAGE_NAME}" ECRDeployment( self, "RoboticsImageUri", src=DockerImageName(local_image.image_uri), dest=DockerImageName(image_uri), ) return image_uri def create_sqs(self, sqs_name: str, env: Environment) -> str: sqs.Queue(self, sqs_name, visibility_timeout=Duration.seconds(60)) return f"https://sqs.{env.region}.amazonaws.com/{env.account}/{sqs_name}" def get_fsx_static_provisioning_manifest( self, volume_handle: str, mount_point: str, env: Environment, ) -> str: t = Template(open(os.path.join(project_dir, "manifests/fsx_static_provisioning.yaml"), "r").read()) return t.substitute( VOLUME_HANDLE=volume_handle, MOUNT_NAME=mount_point, REGION=env.region, )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/eureka/app.py
modules/simulations/eureka/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from typing import cast from aws_cdk import App, CfnOutput, Environment from stack import EurekaStack # Project specific project_name = os.getenv("SEEDFARMER_PROJECT_NAME") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME") module_name = os.getenv("SEEDFARMER_MODULE_NAME") if len(f"{project_name}-{deployment_name}") > 36: raise ValueError("This module cannot support a project+deployment name character length greater than 35") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" eks_cluster_admin_role_arn = os.getenv(_param("EKS_CLUSTER_ADMIN_ROLE_ARN"), "") eks_cluster_name = os.getenv(_param("EKS_CLUSTER_NAME"), "") eks_oidc_arn = os.getenv(_param("EKS_OIDC_ARN"), "") eks_cluster_open_id_connect_issuer = os.getenv(_param("EKS_CLUSTER_OPEN_ID_CONNECT_ISSUER"), "") application_ecr_name = os.getenv(_param("APPLICATION_ECR_NAME"), "") sqs_name = os.getenv(_param("SQS_NAME"), "") fsx_volume_handle = os.getenv(_param("FSX_VOLUME_HANDLE"), "") fsx_mount_point = os.getenv(_param("FSX_MOUNT_POINT"), "") data_bucket_name = os.getenv(_param("DATA_BUCKET_NAME"), "") def generate_description() -> str: soln_id = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_ID", None) soln_name = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_NAME", None) soln_version = os.getenv("SEEDFARMER_PARAMETER_SOLUTION_VERSION", None) desc = "My Module Default Description" if soln_id and soln_name and soln_version: desc = f"({soln_id}) {soln_name}. Version {soln_version}" elif soln_id and soln_name: desc = f"({soln_id}) {soln_name}" return desc app = App() eureka_stack = EurekaStack( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=cast(str, project_name), deployment_name=cast(str, deployment_name), module_name=cast(str, module_name), stack_description=generate_description(), eks_cluster_name=eks_cluster_name, eks_cluster_admin_role_arn=eks_cluster_admin_role_arn, eks_oidc_arn=eks_oidc_arn, eks_cluster_open_id_connect_issuer=eks_cluster_open_id_connect_issuer, simulation_data_bucket_name=data_bucket_name, sqs_name=sqs_name, fsx_volume_handle=fsx_volume_handle, fsx_mount_point=fsx_mount_point, application_ecr_name=application_ecr_name, env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=eureka_stack, id="metadata", value=eureka_stack.to_json_string( { "IamRoleArn": eureka_stack.iam_role_arn, "ApplicationImageUri": eureka_stack.application_image_uri, "SqsUrl": eureka_stack.sqs_url, } ), ) app.synth()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/eureka/tests/test_app.py
modules/simulations/eureka/tests/test_app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import pytest @pytest.fixture(scope="function") def stack_defaults(): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project" os.environ["SEEDFARMER_DEPLOYMENT_NAME"] = "test-deployment" os.environ["SEEDFARMER_MODULE_NAME"] = "test-module" os.environ["CDK_DEFAULT_ACCOUNT"] = "12345678" os.environ["CDK_DEFAULT_REGION"] = "us-east-2" os.environ["SEEDFARMER_PARAMETER_EKS_CLUSTER_ADMIN_ROLE_ARN"] = "arn:aws:iam:us-east-1:1234567890:role/test-role" os.environ["SEEDFARMER_PARAMETER_EKS_CLUSTER_NAME"] = "test_cluster" os.environ["SEEDFARMER_PARAMETER_EKS_OIDC_ARN"] = ( "arn:aws:eks:us-east-1:1234567890:oidc-provider/oidc-provider/oidc.eks.us-east-1.amazonaws.com/id/test-ocid" ) os.environ["SEEDFARMER_PARAMETER_EKS_CLUSTER_OPEN_ID_CONNECT_ISSUER"] = "test_open_id_connect_issuer" os.environ["SEEDFARMER_PARAMETER_APPLICATION_ECR_URI"] = "1234567890.dkr.ecr.us-east-1.amazonaws.com/test-repo" os.environ["SEEDFARMER_PARAMETER_SQS_NAME"] = "test-message-queue" os.environ["SEEDFARMER_PARAMETER_FSX_VOLUME_HANDLE"] = "fs-12345678" os.environ["SEEDFARMER_PARAMETER_FSX_MOUNT_POINT"] = "mnttest" os.environ["SEEDFARMER_PARAMETER_DATA_BUCKET_NAME"] = "test-data-bucket" if "app" in sys.modules: del sys.modules["app"] def test_app(stack_defaults): import app # noqa: F401 def test_project_deployment_name_length(stack_defaults): os.environ["SEEDFARMER_PROJECT_NAME"] = "test-project-incredibly" with pytest.raises(Exception) as e: import app # noqa: F401 assert "module cannot support a project+deployment name character length greater than" in str(e)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/eureka/tests/test_stack.py
modules/simulations/eureka/tests/test_stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os import sys import aws_cdk as cdk import pytest from aws_cdk import Environment from aws_cdk.assertions import Template @pytest.fixture(scope="function") def stack_defaults(): os.environ["CDK_DEFAULT_ACCOUNT"] = "1234567890" os.environ["CDK_DEFAULT_REGION"] = "us-east-1" # Unload the app import so that subsequent tests don't reuse if "stack" in sys.modules: del sys.modules["stack"] def test_app(stack_defaults): import stack app = cdk.App() stack = stack.EurekaStack( scope=app, id="test-proj", project_name="test_proj", deployment_name="test_deploy", module_name="test_module", stack_description="this_is_test_stack", eks_cluster_name="test_cluster", eks_cluster_admin_role_arn="arn:aws:iam:us-east-1:1234567890:role/test-role", eks_oidc_arn="arn:aws:eks:us-east-1:1234567890:oidc-provider/oidc-provider/oidc.eks.us-east-1.amazonaws.com/id/test-ocid", eks_cluster_open_id_connect_issuer="test_open_id_connect_issuer", simulation_data_bucket_name="test-bucket", sqs_name="message-queue", fsx_volume_handle="fs-12345678", fsx_mount_point="mntmount", application_ecr_name="docker.ecr.test_image_name", env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) template = Template.from_stack(stack) template.resource_count_is("AWS::IAM::Role", 2) template.resource_count_is("AWS::IAM::Policy", 2) template.resource_count_is("AWS::SQS::Queue", 1) template.resource_count_is("Custom::CDKBucketDeployment", 1) template.resource_count_is("Custom::AWSCDK-EKS-KubernetesResource", 2)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/eureka/tests/__init__.py
modules/simulations/eureka/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/stack.py
modules/simulations/k8s-managed/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, cast import cdk_nag from aws_cdk import Aspects, Stack, Tags, aws_eks, aws_iam from aws_cdk.lambda_layer_kubectl_v29 import KubectlV29Layer from cdk_nag import NagPackSuppression, NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class SimulationDags(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, mwaa_exec_role: str, eks_cluster_name: str, eks_admin_role_arn: str, eks_openid_connect_provider_arn: str, **kwargs: Any, ) -> None: # Env vars self.project_name = project_name self.deployment_name = deployment_name self.module_name = module_name self.mwaa_exec_role = mwaa_exec_role super().__init__( scope, id, description="(SO9154) Autonomous Driving Data Framework (ADDF) - k8s-managed", **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") # Create Dag IAM Role and policy policy_statements = [ aws_iam.PolicyStatement( actions=["sqs:*"], effect=aws_iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:sqs:{self.region}:{self.account}:{project_name}-{deployment_name}-{module_name}*" ], ), aws_iam.PolicyStatement( actions=["ecr:*"], effect=aws_iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:ecr:{self.region}:{self.account}:repository/{project_name}-{deployment_name}-{module_name}*" ], ), ] dag_document = aws_iam.PolicyDocument(statements=policy_statements) r_name = f"{self.project_name}-{self.deployment_name}-{self.module_name}-dag-role" self.dag_role = aws_iam.Role( self, f"dag-role-{self.deployment_name}-{self.module_name}", assumed_by=aws_iam.ArnPrincipal(self.mwaa_exec_role), inline_policies={"DagPolicyDocument": dag_document}, role_name=r_name, path="/", ) provider = aws_eks.OpenIdConnectProvider.from_open_id_connect_provider_arn( self, "Provider", eks_openid_connect_provider_arn ) cluster = aws_eks.Cluster.from_cluster_attributes( self, f"eks-{self.deployment_name}-{self.module_name}", cluster_name=eks_cluster_name, open_id_connect_provider=provider, kubectl_role_arn=eks_admin_role_arn, kubectl_layer=KubectlV29Layer(self, "Kubectlv29Layer"), ) namespace = cluster.add_manifest( "namespace", { "apiVersion": "v1", "kind": "Namespace", "metadata": {"name": module_name}, }, ) service_account = cluster.add_service_account("service-account", name=module_name, namespace=module_name) service_account.node.add_dependency(namespace) service_account_role: aws_iam.Role = cast(aws_iam.Role, service_account.role) if service_account_role.assume_role_policy: service_account_role.assume_role_policy.add_statements( aws_iam.PolicyStatement( effect=aws_iam.Effect.ALLOW, actions=["sts:AssumeRole"], principals=[aws_iam.ArnPrincipal(mwaa_exec_role)], ) ) for statement in policy_statements: service_account_role.add_to_policy(statement=statement) rbac_role = cluster.add_manifest( "rbac-role", { "apiVersion": "rbac.authorization.k8s.io/v1", "kind": "Role", "metadata": {"name": "module-owner", "namespace": module_name}, "rules": [{"apiGroups": ["*"], "resources": ["*"], "verbs": ["*"]}], }, ) rbac_role.node.add_dependency(namespace) rbac_role_binding = cluster.add_manifest( "rbac-role-binding", { "apiVersion": "rbac.authorization.k8s.io/v1", "kind": "RoleBinding", "metadata": {"name": module_name, "namespace": module_name}, "roleRef": { "apiGroup": "rbac.authorization.k8s.io", "kind": "Role", "name": "module-owner", }, "subjects": [ {"kind": "User", "name": f"{project_name}-{module_name}"}, { "kind": "ServiceAccount", "name": module_name, "namespace": module_name, }, ], }, ) rbac_role_binding.node.add_dependency(service_account) rbac_role = cluster.add_manifest( "rbac-role-default", { "apiVersion": "rbac.authorization.k8s.io/v1", "kind": "Role", "metadata": {"name": "default-access", "namespace": "default"}, "rules": [ { "apiGroups": ["*"], "resources": ["*"], "verbs": ["get", "list", "watch"], } ], }, ) rbac_role.node.add_dependency(namespace) rbac_role_binding = cluster.add_manifest( "rbac-role-binding-default", { "apiVersion": "rbac.authorization.k8s.io/v1", "kind": "RoleBinding", "metadata": {"name": "default-access", "namespace": "default"}, "roleRef": { "apiGroup": "rbac.authorization.k8s.io", "kind": "Role", "name": "default-access", }, "subjects": [ {"kind": "User", "name": f"{project_name}-{module_name}"}, { "kind": "ServiceAccount", "name": module_name, "namespace": module_name, }, ], }, ) rbac_role_binding.node.add_dependency(service_account) rbac_cluster_role_binding = cluster.add_manifest( "rbac-cluster-role-binding", { "apiVersion": "rbac.authorization.k8s.io/v1", "kind": "ClusterRoleBinding", "metadata": {"name": f"system-access-{module_name}"}, "roleRef": { "apiGroup": "rbac.authorization.k8s.io", "kind": "ClusterRole", "name": "system-access", }, "subjects": [ {"kind": "User", "name": f"{project_name}-{module_name}"}, { "kind": "ServiceAccount", "name": module_name, "namespace": module_name, }, ], }, ) rbac_cluster_role_binding.node.add_dependency(service_account) self.eks_service_account_role = service_account.role Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) NagSuppressions.add_stack_suppressions( self, apply_to_nested_stacks=True, suppressions=[ NagPackSuppression( **{ "id": "AwsSolutions-IAM4", "reason": "Managed Policies are for service account roles only", } ), NagPackSuppression( **{ "id": "AwsSolutions-IAM5", "reason": "Resource access restriced to ADDF resources", } ), NagPackSuppression( **{ "id": "AwsSolutions-L1", "reason": "Using latest kubectl v1.29", } ), ], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/app.py
modules/simulations/k8s-managed/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from aws_cdk import App, CfnOutput, Environment from stack import SimulationDags project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") mwaa_exec_role = os.getenv("SEEDFARMER_PARAMETER_MWAA_EXEC_ROLE", "") eks_cluster_name = os.getenv("SEEDFARMER_PARAMETER_EKS_CLUSTER_NAME", "") eks_admin_role_arn = os.getenv("SEEDFARMER_PARAMETER_EKS_CLUSTER_ADMIN_ROLE_ARN", "") eks_oidc_provider_arn = os.getenv("SEEDFARMER_PARAMETER_EKS_OIDC_ARN", "") app = App() stack = SimulationDags( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, mwaa_exec_role=mwaa_exec_role, eks_cluster_name=eks_cluster_name, eks_admin_role_arn=eks_admin_role_arn, eks_openid_connect_provider_arn=eks_oidc_provider_arn, env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "DagRoleArn": stack.dag_role.role_arn, "EksServiceAccountRoleArn": stack.eks_service_account_role.role_arn, } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/tests/test_stack.py
modules/simulations/k8s-managed/tests/test_stack.py
def test_placeholder() -> None: return None
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/tests/__init__.py
modules/simulations/k8s-managed/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/fine_parallel_mock.py
modules/simulations/k8s-managed/simulation_dags/fine_parallel_mock.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging import os import random import string from datetime import timedelta from typing import Any, Dict, Iterator, List, TypeVar import boto3 from airflow import DAG from airflow.models.taskinstance import TaskInstance from airflow.operators.python import PythonOperator from airflow.utils.dates import days_ago from boto3.session import Session from mypy_boto3_sqs.client import SQSClient from simulation_dags import dag_config from simulation_dags.eks_job_operator import EksJobOperator ValueType = TypeVar("ValueType") DAG_ID = os.path.basename(__file__).replace(".py", "") DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("WARNING") def get_client() -> SQSClient: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session.client("sqs") def create_and_populate_queue(num_items: int) -> str: client = get_client() response = client.create_queue( QueueName=( f"{dag_config.PROJECT_NAME}-{dag_config.DEPLOYMENT_NAME}-{dag_config.MODULE_NAME}-" f"{''.join(random.choices(string.ascii_letters + string.digits, k=6))}" ) ) queue_url = response["QueueUrl"] def chunks(lst: List[ValueType], n: int) -> Iterator[ValueType]: """Yield successive n-sized chunks from lst.""" for i in range(0, len(lst), n): yield lst[i : i + n] for i, chunk in enumerate(chunks(range(num_items), 10)): entries = [{"Id": str(j), "MessageBody": f"message-{i}-{j}-{value}"} for j, value in enumerate(chunk)] client.send_message_batch(QueueUrl=queue_url, Entries=entries) return queue_url def delete_queue(ti: TaskInstance, create_queue_task_id: str) -> None: client = get_client() url = ti.xcom_pull(task_ids=create_queue_task_id) client.delete_queue(QueueUrl=url) def get_job_body(create_queue_task_id: str, parallelism: int) -> Dict[str, Any]: return { "apiVerson": "batch/v1", "kind": "Job", "metadata": {}, "spec": { "parallelism": parallelism, "completions": parallelism, "backoffLimit": 6, "template": { "spec": { "serviceAccountName": dag_config.MODULE_NAME, "restartPolicy": "OnFailure", "volumes": [{"name": "shared-data", "emptyDir": {}}], "containers": [ { "name": "sqs-manager", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/sqs_manager.py", ], "args": ["--url", "$(URL)", "--dir", "$(DIR)"], "env": [ { "name": "URL", "value": f"{{{{ ti.xcom_pull(task_ids='{create_queue_task_id}') }}}}", }, {"name": "DIR", "value": "/shared-data"}, {"name": "DEBUG", "value": "true"}, { "name": "AWS_DEFAULT_REGION", "value": dag_config.REGION, }, { "name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID, }, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, { "name": "simulator", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/simulator.py", ], "args": [ "--dir", "$(DIR)", "--max-seconds", "$(MAX_SECONDS)", "--failure-seed", "$(FAILURE_SEED)", ], "env": [ {"name": "DIR", "value": "/shared-data"}, {"name": "MAX_SECONDS", "value": "60"}, # probability of failure = 1/FAILURE_SEED if 1 <= FAILURE_SEED <= 32768 else 0 {"name": "FAILURE_SEED", "value": "32769"}, {"name": "DEBUG", "value": "true"}, { "name": "AWS_DEFAULT_REGION", "value": dag_config.REGION, }, { "name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID, }, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, ], } }, }, } with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=6), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: total_simulations = 50 parallelism = 10 create_queue_task = PythonOperator( task_id="create-queue", dag=dag, provide_context=True, python_callable=create_and_populate_queue, op_kwargs={"num_items": total_simulations}, ) job_task = EksJobOperator( task_id="fine-parallel-mock", dag=dag, namespace=dag_config.EKS_NAMESPACE, # type: ignore body=get_job_body( create_queue_task_id=create_queue_task.task_id, parallelism=parallelism, ), delete_policy="IfSucceeded", cluster_name=dag_config.EKS_CLUSTER_NAME, # type: ignore service_account_role_arn=dag_config.EKS_SERVICE_ACCOUNT_ROLE, # type: ignore get_logs=False, ) delete_queue_task = PythonOperator( task_id="delete_queue", dag=dag, provide_context=True, python_callable=delete_queue, op_kwargs={"create_queue_task_id": create_queue_task.task_id}, ) create_queue_task >> job_task >> delete_queue_task
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/coarse_parallel_mock.py
modules/simulations/k8s-managed/simulation_dags/coarse_parallel_mock.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging import os import random import string from datetime import timedelta from typing import Any, Dict, Iterator, List, TypeVar import boto3 from airflow import DAG from airflow.models.taskinstance import TaskInstance from airflow.operators.python import PythonOperator from airflow.utils.dates import days_ago from boto3.session import Session from mypy_boto3_sqs.client import SQSClient from simulation_dags import dag_config from simulation_dags.eks_job_operator import EksJobOperator ValueType = TypeVar("ValueType") DAG_ID = os.path.basename(__file__).replace(".py", "") DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("WARNING") def get_client() -> SQSClient: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session.client("sqs") def create_and_populate_queue(num_items: int) -> str: client = get_client() response = client.create_queue( QueueName=( f"{dag_config.PROJECT_NAME}-{dag_config.DEPLOYMENT_NAME}-{dag_config.MODULE_NAME}-" f"{''.join(random.choices(string.ascii_letters + string.digits, k=6))}" ) ) queue_url = response["QueueUrl"] def chunks(lst: List[ValueType], n: int) -> Iterator[ValueType]: """Yield successive n-sized chunks from lst.""" for i in range(0, len(lst), n): yield lst[i : i + n] for i, chunk in enumerate(chunks(range(num_items), 10)): entries = [{"Id": str(j), "MessageBody": f"message-{i}-{j}-{value}"} for j, value in enumerate(chunk)] client.send_message_batch(QueueUrl=queue_url, Entries=entries) return queue_url def delete_queue(ti: TaskInstance, create_queue_task_id: str) -> None: client = get_client() url = ti.xcom_pull(task_ids=create_queue_task_id) client.delete_queue(QueueUrl=url) def get_job_body(create_queue_task_id: str, parallelism: int, completions: int) -> Dict[str, Any]: return { "apiVerson": "batch/v1", "kind": "Job", "metadata": {}, "spec": { "parallelism": parallelism, "completions": completions, "backoffLimit": 6, "template": { "spec": { "serviceAccountName": dag_config.MODULE_NAME, "restartPolicy": "Never", "volumes": [{"name": "shared-data", "emptyDir": {}}], "containers": [ { "name": "sqs-manager", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/sqs_manager.py", ], "args": [ "--url", "$(URL)", "--dir", "$(DIR)", "--single-message", ], "env": [ { "name": "URL", "value": f"{{{{ ti.xcom_pull(task_ids='{create_queue_task_id}') }}}}", }, {"name": "DIR", "value": "/shared-data"}, {"name": "DEBUG", "value": "true"}, { "name": "AWS_DEFAULT_REGION", "value": dag_config.REGION, }, { "name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID, }, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, { "name": "simulator", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/simulator.py", ], "args": [ "--dir", "$(DIR)", "--max-seconds", "$(MAX_SECONDS)", "--failure-seed", "$(FAILURE_SEED)", ], "env": [ {"name": "DIR", "value": "/shared-data"}, {"name": "MAX_SECONDS", "value": "60"}, # probability of failure = 1/FAILURE_SEED if 1 <= FAILURE_SEED <= 32768 else 0 {"name": "FAILURE_SEED", "value": "32769"}, {"name": "DEBUG", "value": "true"}, { "name": "AWS_DEFAULT_REGION", "value": dag_config.REGION, }, { "name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID, }, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, ], } }, }, } with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=6), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: total_simulations = 50 parallelism = 10 create_queue_task = PythonOperator( task_id="create-queue", dag=dag, provide_context=True, python_callable=create_and_populate_queue, op_kwargs={"num_items": total_simulations}, ) job_task = EksJobOperator( task_id="coarse-parallel-mock", dag=dag, namespace=dag_config.EKS_NAMESPACE, # type: ignore body=get_job_body( create_queue_task_id=create_queue_task.task_id, parallelism=parallelism, completions=total_simulations, ), delete_policy="IfSucceeded", cluster_name=dag_config.EKS_CLUSTER_NAME, # type: ignore service_account_role_arn=dag_config.EKS_SERVICE_ACCOUNT_ROLE, # type: ignore get_logs=False, ) delete_queue_task = PythonOperator( task_id="delete_queue", dag=dag, provide_context=True, python_callable=delete_queue, op_kwargs={"create_queue_task_id": create_queue_task.task_id}, ) create_queue_task >> job_task >> delete_queue_task
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/dag_config.py
modules/simulations/k8s-managed/simulation_dags/dag_config.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0. # This file is populated with configurations information when the Module is deployed # Configuration parameters are exported as module level constants # # Example: # SOME_PARAMETER = 'some value'
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/eks_job_operator.py
modules/simulations/k8s-managed/simulation_dags/eks_job_operator.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging import subprocess import tempfile from threading import Timer from typing import Any from airflow_kubernetes_job_operator import KubernetesJobOperator logging.basicConfig(level="DEBUG") logger = logging.getLogger("airflow") KUBECONFIG_REFRESH_RATE = 180 class EksJobOperator(KubernetesJobOperator): # type: ignore def __init__( self, *args: Any, cluster_name: str, service_account_role_arn: str, **kwargs: Any, ) -> None: self.cluster_name = cluster_name self.service_account_role_arn = service_account_role_arn self.logger = logger super().__init__(*args, **kwargs) def update_kubeconfig(self, path: str) -> None: args = [ "aws", "eks", "update-kubeconfig", "--name", self.cluster_name, "--role-arn", self.service_account_role_arn, "--kubeconfig", path, ] logger.info("command: %s", args) subprocess.check_call(args) def load_kubeconfig(self) -> None: logger.debug("Reloading kubeconfig") # Load the configuration. self.job_runner.client.load_kube_config( config_file=self.config_file, is_in_cluster=self.in_cluster, context=self.cluster_context, ) Timer(KUBECONFIG_REFRESH_RATE, self.load_kubeconfig).start() def pre_execute(self, context: Any) -> Any: """Called before execution by the airflow system. Overriding this method without calling its super() will break the job operator. Arguments: context -- The airflow context """ # Load the configuration. with tempfile.NamedTemporaryFile() as temp: kubeconfig_path = temp.name self.update_kubeconfig(path=kubeconfig_path) self.config_file = kubeconfig_path self.in_cluster = False self.cluster_context = None # This background thread refereshes the kubeconfig to eliminate # timeouts for long running jobs Timer(KUBECONFIG_REFRESH_RATE, self.load_kubeconfig).start() # call parent. return super().pre_execute(context)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/coarse_fan_out_mock.py
modules/simulations/k8s-managed/simulation_dags/coarse_fan_out_mock.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import json import logging import os import random import string from datetime import timedelta from typing import Any, Dict, Iterator, List, TypeVar import boto3 from airflow import DAG from airflow.exceptions import AirflowFailException from airflow.models.taskinstance import TaskInstance from airflow.operators.python import PythonOperator from airflow.utils.dates import days_ago from boto3.session import Session from mypy_boto3_sqs.client import SQSClient from simulation_dags import dag_config from simulation_dags.eks_job_operator import EksJobOperator ValueType = TypeVar("ValueType") DAG_ID = os.path.basename(__file__).replace(".py", "") DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("DEBUG") def get_client() -> SQSClient: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session.client("sqs") def create_and_populate_queue(num_items: int) -> str: client = get_client() response = client.create_queue( QueueName=( f"{dag_config.PROJECT_NAME}-{dag_config.DEPLOYMENT_NAME}-{dag_config.MODULE_NAME}-" f"{''.join(random.choices(string.ascii_letters + string.digits, k=6))}" ) ) queue_url = response["QueueUrl"] def chunks(lst: List[ValueType], n: int) -> Iterator[ValueType]: """Yield successive n-sized chunks from lst.""" for i in range(0, len(lst), n): yield lst[i : i + n] for i, chunk in enumerate(chunks(range(num_items), 10)): entries = [{"Id": str(j), "MessageBody": f"message-{i}-{j}-{value}"} for j, value in enumerate(chunk)] client.send_message_batch(QueueUrl=queue_url, Entries=entries) return queue_url def fail_if_job_failed(ti: TaskInstance, eks_job_task_id: str) -> None: status = ti.xcom_pull(task_ids=eks_job_task_id, key="status") print(f"eks_job_task_id: {eks_job_task_id} status: {status}") if status != "JOB_COMPLETED": raise AirflowFailException("EKS Job Failed") def delete_queue(ti: TaskInstance, create_queue_task_id: str) -> None: client = get_client() url = ti.xcom_pull(task_ids=create_queue_task_id) client.delete_queue(QueueUrl=url) def get_job_body(create_queue_task_id: str, parallelism: int, completions: int, max_failures: int) -> Dict[str, Any]: worker_pod_body = json.dumps( { "apiVersion": "v1", "kind": "Pod", "metadata": {}, "spec": { "serviceAccountName": dag_config.MODULE_NAME, "restartPolicy": "Never", "volumes": [{"name": "shared-data", "emptyDir": {}}], "containers": [ { "name": "sqs-manager", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/sqs_manager.py", ], "args": [ "--url", "$(QUEUE_URL)", "--dir", "$(DIR)", "--single-message", ], "env": [ {"name": "DIR", "value": "/shared-data"}, {"name": "DEBUG", "value": "true"}, {"name": "AWS_DEFAULT_REGION", "value": dag_config.REGION}, {"name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID}, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, { "name": "simulator", "image": dag_config.SIMULATION_MOCK_IMAGE, "volumeMounts": [{"name": "shared-data", "mountPath": "/shared-data"}], "command": [ "python", "/var/simulation-mock/simulation_mock/simulator.py", ], "args": [ "--dir", "$(DIR)", "--max-seconds", "$(MAX_SECONDS)", "--failure-seed", "$(FAILURE_SEED)", ], "env": [ {"name": "DIR", "value": "/shared-data"}, {"name": "MAX_SECONDS", "value": "60"}, # probability of failure = 1/FAILURE_SEED if 1 <= FAILURE_SEED <= 32768 else 0 {"name": "FAILURE_SEED", "value": "32769"}, {"name": "DEBUG", "value": "true"}, {"name": "AWS_DEFAULT_REGION", "value": dag_config.REGION}, {"name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID}, ], "livenessProbe": { "exec": { "command": ["cat", "/tmp/container.running"], }, "initialDelaySeconds": 5, "periodSeconds": 5, }, }, ], }, } ) return { "apiVerson": "batch/v1", "kind": "Job", "metadata": { "labels": {"app": "addf-job"}, }, "spec": { "parallelism": 1, "completions": 1, "backoffLimit": 6, "template": { "metadata": { "labels": {"app": "addf-job"}, }, "spec": { "serviceAccountName": dag_config.MODULE_NAME, "restartPolicy": "Never", "priorityClassName": "system-node-critical", "containers": [ { "name": "pod-launcher", "image": dag_config.SIMULATION_MOCK_IMAGE, "command": ["kopf", "run"], "args": [ "simulation_mock/pod_launcher.py", "--namespace", dag_config.EKS_NAMESPACE, # "--verbose" ], "resources": { "requests": {"memory": "512Mi", "cpu": "1"}, "limits": {"memory": "1Gi", "cpu": "1"}, }, "env": [ { "name": "NAMESPACE", "value": dag_config.EKS_NAMESPACE, }, { "name": "POD_LAUNCHER_NAME", "valueFrom": {"fieldRef": {"fieldPath": "metadata.name"}}, }, { "name": "POD_LAUNCHER_UID", "valueFrom": {"fieldRef": {"fieldPath": "metadata.uid"}}, }, { "name": "JOB_NAME", "valueFrom": {"fieldRef": {"fieldPath": "metadata.labels['job-name']"}}, }, { "name": "JOB_UID", "valueFrom": {"fieldRef": {"fieldPath": "metadata.labels['controller-uid']"}}, }, {"name": "PARALLELISM", "value": f"{parallelism}"}, {"name": "COMPLETIONS", "value": f"{completions}"}, {"name": "MAX_FAILURES", "value": f"{max_failures}"}, { "name": "AWS_DEFAULT_REGION", "value": dag_config.REGION, }, { "name": "AWS_ACCOUNT_ID", "value": dag_config.ACCOUNT_ID, }, { "name": "QUEUE_URL", "value": f"{{{{ ti.xcom_pull(task_ids='{create_queue_task_id}') }}}}", }, {"name": "WORKER_POD_BODY", "value": worker_pod_body}, ], }, ], }, }, }, } with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=6), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: total_simulations = 50 parallelism = 10 max_failures = 5 create_queue_task = PythonOperator( task_id="create-queue", dag=dag, provide_context=True, python_callable=create_and_populate_queue, op_kwargs={"num_items": total_simulations}, ) job_task = EksJobOperator( task_id="coarse-fan-out-mock", dag=dag, namespace=dag_config.EKS_NAMESPACE, # type: ignore body=get_job_body( create_queue_task_id=create_queue_task.task_id, parallelism=parallelism, completions=total_simulations, max_failures=max_failures, ), delete_policy="Never", cluster_name=dag_config.EKS_CLUSTER_NAME, # type: ignore service_account_role_arn=dag_config.EKS_SERVICE_ACCOUNT_ROLE, # type: ignore get_logs=True, ) fail_if_job_failed_task = PythonOperator( task_id="fail-if-job-failed", dag=dag, provide_context=True, python_callable=fail_if_job_failed, op_kwargs={"eks_job_task_id": job_task.task_id}, ) delete_queue_task = PythonOperator( task_id="delete-queue", dag=dag, provide_context=True, python_callable=delete_queue, op_kwargs={"create_queue_task_id": create_queue_task.task_id}, ) create_queue_task >> job_task >> fail_if_job_failed_task >> delete_queue_task
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/utils.py
modules/simulations/k8s-managed/simulation_dags/utils.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import boto3 from boto3.session import Session def get_assumed_role_session(role_arn: str) -> Session: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=role_arn, RoleSessionName="AssumeRoleSession1", ) return Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/__init__.py
modules/simulations/k8s-managed/simulation_dags/__init__.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License.
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/simulation_dags/simple_mock.py
modules/simulations/k8s-managed/simulation_dags/simple_mock.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging import os import textwrap from datetime import timedelta from airflow import DAG from airflow.utils.dates import days_ago from simulation_dags import dag_config from simulation_dags.eks_job_operator import EksJobOperator DAG_ID = os.path.basename(__file__).replace(".py", "") DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("DEBUG") with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: # caller_identity = PythonOperator(task_id="log_caller_identity", dag=dag, python_callable=log_caller_identity) envs = {"JOB_NAME": "Simple Mock", "MAX_SECONDS": "120", "FAILURE_SEED": "20"} default_delete_policy = "IfSucceeded" body = { "apiVerson": "batch/v1", "kind": "Job", "metadata": {}, "spec": { "parallelism": 10, "completions": 50, "backoffLimit": 6, "template": { "spec": { "restartPolicy": "OnFailure", "containers": [ { "name": "job-executor", "image": "ubuntu", "command": [ "bash", "-c", textwrap.dedent( """\ #/usr/bin/env bash echo "[$(date)] Starting $JOB_NAME" TIC_COUNT=$RANDOM let "TIC_COUNT %= $MAX_SECONDS" FAILURE_CHECK=$RANDOM let "FAILURE_CHECK %= $FAILURE_SEED" echo "[$(date)] Random Runtime: $TIC_COUNT" echo "[$(date)] Random Failure: $FAILURE_CHECK" CURRENT_COUNT=0 while true; do CURRENT_COUNT=$((CURRENT_COUNT + 1)) if [ "$CURRENT_COUNT" -ge "$TIC_COUNT" ]; then break fi sleep 1 done if [ "$FAILURE_CHECK" -eq "0" ]; then echo "[$(date)] Failure" && exit 1 else echo "[$(date)] Complete" && exit 0 fi """ ), ], } ], } }, }, } job_task = EksJobOperator( task_id="bash-mock-job", dag=dag, namespace=dag_config.EKS_NAMESPACE, # type: ignore envs=envs, body=body, delete_policy=default_delete_policy, cluster_name=dag_config.EKS_CLUSTER_NAME, # type: ignore service_account_role_arn=dag_config.EKS_SERVICE_ACCOUNT_ROLE, # type: ignore ) job_task
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/setup.py
modules/simulations/k8s-managed/images/simulation-mock/src/setup.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from setuptools import find_packages, setup setup( name="simulation-mock", version="0.1.0", author="AWS Professional Services", author_email="aws-proserve-opensource@amazon.com", project_urls={"Org Site": "https://aws.amazon.com/professional-services/"}, packages=find_packages(include=["simulation_mock", "simulation_mock.*"]), python_requires=">=3.7, <3.14", install_requires=[ "boto3~=1.21.19", "platonic-sqs==1.0.1", "kubernetes~=21.7.0", "kopf~=1.35.4", ], include_package_data=True, )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/pod_launcher.py
modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/pod_launcher.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import json import logging import os import subprocess from typing import Any, Dict import kopf from kubernetes import config as k8_config from kubernetes import dynamic from kubernetes.client import api_client NAMESPACE = os.environ.get("NAMESPACE") POD_LAUNCHER_NAME = os.environ.get("POD_LAUNCHER_NAME") POD_LAUNCHER_UID = os.environ.get("POD_LAUNCHER_UID") JOB_NAME = os.environ.get("JOB_NAME") JOB_UID = os.environ.get("JOB_UID") POD_LAUNCHER_BODY = None COMPLETIONS = 0 FAILURES = 0 def load_config(in_cluster: bool = True) -> None: in_cluster_env = os.environ.get("IN_CLUSTER_DEPLOYMENT", None) in_cluster = in_cluster_env.lower() in ["yes", "true", "1"] if in_cluster_env is not None else in_cluster if in_cluster: k8_config.load_incluster_config() else: k8_config.load_kube_config() def dynamic_client() -> dynamic.DynamicClient: load_config() return dynamic.DynamicClient(client=api_client.ApiClient()) @kopf.on.startup() def configure(settings: kopf.OperatorSettings, memo: kopf.Memo, logger: kopf.Logger, **_: Any) -> None: settings.persistence.progress_storage = kopf.MultiProgressStorage( [ kopf.AnnotationsProgressStorage(prefix="simulation-pod-launcher.addf.aws"), kopf.StatusProgressStorage(field="simulation-pod-launcher.addf.aws"), ] ) settings.persistence.finalizer = "simulation-pod-launcher.addf.aws/kopf-finalizer" settings.posting.level = logging.getLevelName(os.environ.get("EVENT_LOG_LEVEL", "INFO")) memo.namespace = os.environ.get("NAMESPACE") memo.pod_launcher_name = os.environ.get("POD_LAUNCHER_NAME") memo.pod_launcher_uid = os.environ.get("POD_LAUNCHER_UID") memo.worker_body = os.environ.get("WORKER_POD_BODY") memo.queue_url = os.environ.get("QUEUE_URL") memo.default_region = os.environ.get("AWS_DEFAULT_REGION") memo.account_id = os.environ.get("AWS_ACCOUNT_ID") memo.parallelism = int(os.environ.get("PARALLELISM")) memo.completions = int(os.environ.get("COMPLETIONS")) memo.max_failures = int(os.environ.get("MAX_FAILURES")) logger.info("memo.namespace: %s", memo.namespace) logger.info("memo.pod_launcher_name: %s", memo.pod_launcher_name) logger.info("memo.pod_launcher_uid: %s", memo.pod_launcher_uid) logger.info("memo.worker_body: %s", memo.worker_body) logger.info("memo.queue_url: %s", memo.queue_url) logger.info("memo.default_region: %s", memo.default_region) logger.info("memo.account_id: %s", memo.account_id) logger.info("memo.parallelism: %s", memo.parallelism) logger.info("JOB_NAME: %s", JOB_NAME) logger.info("JOB_UID: %s", JOB_UID) def _should_index_pod(meta: kopf.Meta, **_: Any) -> bool: return ( "ownerReferences" in meta and len(meta["ownerReferences"]) > 0 and meta["ownerReferences"][0]["uid"] == POD_LAUNCHER_UID ) @kopf.index("pods", when=_should_index_pod) def worker_pods_idx(name: str, status: kopf.Status, **_: Any) -> Dict[str, str]: status = status.get("phase") if status not in ["Succeeded", "Failed"]: return {"Running": name} elif status == "Succeeded": global COMPLETIONS COMPLETIONS += 1 return {"Completed": name} elif status == "Failed": global FAILURES FAILURES += 1 return {"Failed": name} else: return {"Unknown": name} def _should_update_job_metadata(meta: kopf.Meta, **_: Any) -> bool: return meta["uid"] == POD_LAUNCHER_UID @kopf.on.resume("pods", when=_should_update_job_metadata) @kopf.on.create("pods", when=_should_update_job_metadata) def update_job_metadata(meta: kopf.Meta, body: kopf.Body, memo: kopf.Memo, logger: kopf.Logger, **_: Any) -> str: global POD_LAUNCHER_BODY POD_LAUNCHER_BODY = body logger.debug("POD_LAUNCHER_BODY: %s", POD_LAUNCHER_BODY) return "METADATA_UPDATED" def _should_monitor_job(meta: kopf.Meta, **_: Any) -> bool: return meta["uid"] == JOB_UID @kopf.daemon("batch", "v1", "jobs", when=_should_monitor_job, initial_delay=30) def monitor_job( name: str, patch: kopf.Patch, stopped: kopf.DaemonStopped, memo: kopf.Memo, logger: kopf.Logger, worker_pods_idx: kopf.Index, **_: Any, ) -> str: logger.info("JOB: %s", name) worker_spec = json.loads(memo.worker_body) for container in worker_spec["spec"]["containers"]: env = container["env"] if "env" in container else [] env.extend( [ {"name": "NAMESPACE", "value": memo.namespace}, {"name": "JOB_NAME", "value": JOB_NAME}, {"name": "JOB_UID", "value": JOB_UID}, {"name": "POD_LAUNCHER_NAME", "value": memo.pod_launcher_name}, {"name": "AWS_DEFAULT_REGION", "value": memo.default_region}, {"name": "AWS_ACCOUNT_ID", "value": memo.account_id}, {"name": "QUEUE_URL", "value": memo.queue_url}, ] ) kopf.append_owner_reference(objs=[worker_spec], owner=POD_LAUNCHER_BODY) kopf.harmonize_naming(objs=[worker_spec], name=POD_LAUNCHER_BODY["metadata"]["name"], forced=True) kopf.adjust_namespace(objs=[worker_spec], namespace=NAMESPACE, forced=True) logger.debug("WORKER_BODY: %s", worker_spec) running_pod_count = 0 while not stopped and COMPLETIONS < memo.completions and FAILURES < memo.max_failures: running_pod_count = len(list(worker_pods_idx.get("Running"))) if "Running" in worker_pods_idx else 0 launch_count = min( memo.parallelism - running_pod_count, memo.completions - (COMPLETIONS + running_pod_count), ) logger.info( "RUNNING_POD_COUNT: %s COMPLETIONS: %s FAILURES: %s LAUNCHING: %s", running_pod_count, COMPLETIONS, FAILURES, launch_count, ) client = dynamic_client() api = client.resources.get(kind="Pod") counter = 0 for _ in range(0, launch_count): counter += 1 try: api.create(namespace=NAMESPACE, body=worker_spec) except Exception as e: logger.exception("LAUNCH_ERROR", e) if counter % 10 == 0: stopped.wait(0.5) stopped.wait(10) if COMPLETIONS < memo.completions and FAILURES >= memo.max_failures: logger.info("FAILED") patch["metadata"] = { "annotations": { "simulation-pod-launcher.addf.aws/status": "Failed", "simulation-pod-launcher.addf.aws/completions": f"{COMPLETIONS}", "simulation-pod-launcher.addf.aws/failures": f"{FAILURES}", } } patch["status"] = {"phase": "Failed"} status = "JOB_FAILED" else: logger.info("COMPLETE") patch["metadata"] = { "annotations": { "simulation-pod-launcher.addf.aws/status": "Succeeded", "simulation-pod-launcher.addf.aws/completions": f"{COMPLETIONS}", "simulation-pod-launcher.addf.aws/failures": f"{FAILURES}", } } status = "JOB_COMPLETED" json_status = json.dumps({"status": status}) logger.info("WILL_EXIT_IN_30") # Write status out so that it can be picked up by the XCom Value Parser # And fork a subprocess to send a SIGTERM and initiate graceful shutdown subprocess.Popen( f"sleep 30; echo '::kube_api:xcom={json_status}'; echo 'Sending SIGTERM'; kill -15 {os.getpid()}", shell=True, ) return status
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/signal_handler.py
modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/signal_handler.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import signal import sys from threading import Event class SignalHandler: kill_now = False def __init__(self, exit: Event): self.exit = exit signal.signal(signal.SIGINT, self.exit_gracefully) signal.signal(signal.SIGTERM, self.exit_gracefully) signal.signal(signal.SIGABRT, self.exit_ungracefully) def exit_gracefully(self, signum, frame): self.kill_now = True self.exit.set() def exit_ungracefully(self, signum, frame): sys.exit(1)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/__init__.py
modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/__init__.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import logging import os DEBUG_LOGGING_FORMAT = "[%(asctime)s][%(filename)-13s:%(lineno)3d][%(levelname)s][%(threadName)s] %(message)s" def get_logger() -> logging.Logger: debug = os.environ.get("DEBUG", "False").lower() in [ "true", "yes", "1", ] level = logging.DEBUG if debug else logging.INFO logging.basicConfig(level=level, format=DEBUG_LOGGING_FORMAT) logger: logging.Logger = logging.getLogger(__name__) logger.setLevel(level) if debug: logging.getLogger("boto3").setLevel(logging.ERROR) logging.getLogger("botocore").setLevel(logging.ERROR) logging.getLogger("urllib3").setLevel(logging.ERROR) return logger
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/sqs_manager.py
modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/sqs_manager.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import argparse import os import sys from contextlib import contextmanager from datetime import timedelta from threading import Event from typing import Iterator from platonic.queue.errors import MessageReceiveTimeout from platonic.sqs.queue import SQSReceiver from platonic.sqs.queue.message import SQSMessage from platonic.sqs.queue.types import ValueType from platonic.timeout import ConstantTimeout from simulation_mock import get_logger from simulation_mock.signal_handler import SignalHandler MESSAGE_RECEIVE_TIMEOUT_SECONDS = 10 PROCESSING_TIMEOUT_MINUTES = 5 FILE_CHECK_SLEEP_TIME_SECONDS = 10 DATA_FILE_NAME = "message.json" LOGGER = get_logger() class ProcessingTimeoutException(Exception): pass class SQSHeartbeatReceiver(SQSReceiver[ValueType]): def heartbeat(self, message: SQSMessage[ValueType], seconds: int) -> None: """Extend the visibility timeout of Messages being processed Parameters ---------- message : SQSMessage[ValueType] The SQS Message to extend seconds : int Number of seconds to extend the timeout """ self.client.change_message_visibility( QueueUrl=self.url, ReceiptHandle=message.receipt_handle, VisibilityTimeout=seconds, ) @contextmanager def acknowledgement(self, message: SQSMessage[ValueType]) -> Iterator[SQSMessage[ValueType]]: """Override of ``acknowledgement`` that won't acknowledge (delete) them message if an Exception is thrown Parameters ---------- message : SQSMessage[ValueType] The SQS Message to conditionally acknowledge Yields ------ SQSMessage[ValueType] The SQS Message Raises ------ e Any exception that was thrown """ try: # noqa: WPS501 yield message except Exception as e: LOGGER.debug("NAK: %s - %s", message, e) raise e else: LOGGER.debug("ACK: %s", message) self.acknowledge(message) def main(url: str, dir: str, single_message: bool) -> int: LOGGER.info("QUEUE: %s", url) LOGGER.info("DIR: %s", dir) LOGGER.info("SINGLE_MESSAGE: %s", single_message) incoming_simulations = SQSHeartbeatReceiver[str]( url=url, timeout=ConstantTimeout(period=timedelta(seconds=MESSAGE_RECEIVE_TIMEOUT_SECONDS)), ) signal_handler = SignalHandler(exit=Event()) while True: try: if signal_handler.kill_now: return 0 with incoming_simulations.acknowledgement(incoming_simulations.receive()) as message: data_file = os.path.join(dir, "message.json") LOGGER.info("MSG: %s", message) LOGGER.info("DATA_FILE: %s", data_file) with open(data_file, "w") as file: file.write(message.value) timeout = ConstantTimeout(timedelta(minutes=PROCESSING_TIMEOUT_MINUTES)) with timeout.timer() as timer: while not timer.is_expired: if signal_handler.kill_now: return 0 if os.path.isfile(data_file): LOGGER.debug("EXISTS: %s", data_file) incoming_simulations.heartbeat(message=message, seconds=(60)) signal_handler.exit.wait(FILE_CHECK_SLEEP_TIME_SECONDS) else: LOGGER.debug("GONE: %s", data_file) break else: raise ProcessingTimeoutException("TIMEOUT") except MessageReceiveTimeout: LOGGER.info("EMPTY: %s", url) return 0 except ProcessingTimeoutException: LOGGER.error("TIMEOUT: %s", PROCESSING_TIMEOUT_MINUTES) if single_message: return 1 else: LOGGER.info("SUCCESS: %s", message) if single_message: return 0 if __name__ == "__main__": parser = argparse.ArgumentParser(description="Manage SQS Messages") parser.add_argument("--url", required=True, help="URL of the SQS Queue to manage") parser.add_argument("--dir", required=True, help="Directory to use/monitor for data files") parser.add_argument( "--single-message", action="store_true", help="Whether to retrieve/process a single message or process messages until the queue is empty", ) args = parser.parse_args() LOGGER.debug("ARGS: %s", args) with open("/tmp/container.running", "w") as file: LOGGER.info("LIVENESS_FILE: /tmp/container.running") sys.exit(main(url=args.url, dir=args.dir, single_message=args.single_message))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/simulator.py
modules/simulations/k8s-managed/images/simulation-mock/src/simulation_mock/simulator.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import argparse import os import sys from datetime import timedelta from random import randint from threading import Event from platonic.timeout import ConstantTimeout from simulation_mock import get_logger from simulation_mock.signal_handler import SignalHandler FILE_CHECK_TIMEOUT_SECONDS = 30 DATA_FILE_NAME = "message.json" FILE_CHECK_SLEEP_TIME_SECONDS = 10 LOGGER = get_logger() class RandomFailureException(Exception): pass class FileCheckTimoutException(Exception): pass def main(dir: str, max_seconds: int, failure_seed: int) -> int: LOGGER.info("DIR: %s", dir) LOGGER.info("MAX_SECONDS: %s", max_seconds) LOGGER.info("FAILURE_SEED: %s", failure_seed) signal_handler = SignalHandler(exit=Event()) while True: try: if signal_handler.kill_now: return 0 timeout = ConstantTimeout(timedelta(seconds=FILE_CHECK_TIMEOUT_SECONDS)) with timeout.timer() as timer: data_file = os.path.join(dir, "message.json") LOGGER.info("DATA_FILE: %s", data_file) while not timer.is_expired: if signal_handler.kill_now: return 0 if os.path.isfile(data_file): LOGGER.info("EXISTS: %s", data_file) sleep_time = randint(0, max_seconds) LOGGER.info("RANDOM_SLEEP: %s", sleep_time) signal_handler.exit.wait(sleep_time) if (randint(1, 32768) % failure_seed) == 0: raise RandomFailureException("RANDOM") else: LOGGER.info("REMOVING: %s", data_file) os.remove(data_file) break else: LOGGER.debug("NOT_EXISTS: %s", data_file) signal_handler.exit.wait(FILE_CHECK_SLEEP_TIME_SECONDS) else: raise FileCheckTimoutException("TIMEOUT") except RandomFailureException: LOGGER.error("RANDOM_FAILURE") return 1 except FileCheckTimoutException: LOGGER.info("EXISTS_TIMEOUT: %s", data_file) return 0 if __name__ == "__main__": parser = argparse.ArgumentParser(description="Mock simulation") parser.add_argument("--dir", required=True, help="Directory to use/monitor for data files") parser.add_argument( "--max-seconds", required=True, type=int, help="Max runtime to execute mock simulation", ) parser.add_argument( "--failure-seed", required=True, type=int, help="Seed number to determine random failures", ) args = parser.parse_args() LOGGER.debug("ARGS: %s", args) with open("/tmp/container.running", "w") as file: LOGGER.info("LIVENESS_FILE: /tmp/container.running") sys.exit(main(dir=args.dir, max_seconds=args.max_seconds, failure_seed=args.failure_seed))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/stack.py
modules/simulations/batch-managed/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import logging from typing import Any, Dict, List, cast import aws_cdk.aws_batch_alpha as batch import aws_cdk.aws_ec2 as ec2 import aws_cdk.aws_iam as iam import cdk_nag from aws_cdk import Aspects, Stack, Tags from cdk_nag import NagSuppressions from constructs import Construct, IConstruct _logger: logging.Logger = logging.getLogger(__name__) class BatchDags(Stack): def __init__( self, scope: Construct, id: str, *, project_name: str, deployment_name: str, module_name: str, mwaa_exec_role: str, vpc_id: str, private_subnet_ids: List[str], batch_compute: Dict[str, Any], **kwargs: Any, ) -> None: # Env vars self.deployment_name = deployment_name self.module_name = module_name self.mwaa_exec_role = mwaa_exec_role super().__init__( scope, id, description="(SO9154) Autonomous Driving Data Framework (ADDF) - batch-managed", **kwargs, ) Tags.of(scope=cast(IConstruct, self)).add(key="Deployment", value=f"{project_name}-{deployment_name}") dep_mod = f"{project_name}-{deployment_name}-{module_name}" self.vpc_id = vpc_id self.vpc = ec2.Vpc.from_lookup( self, "VPC", vpc_id=vpc_id, ) self.private_subnets = [] for idx, subnet_id in enumerate(private_subnet_ids): self.private_subnets.append(ec2.Subnet.from_subnet_id(scope=self, id=f"subnet{idx}", subnet_id=subnet_id)) # Create Dag IAM Role and policy policy_statements = [ iam.PolicyStatement( actions=["sqs:*"], effect=iam.Effect.ALLOW, resources=[f"arn:{self.partition}:sqs:{self.region}:{self.account}:{dep_mod}*"], ), iam.PolicyStatement( actions=["ecr:*"], effect=iam.Effect.ALLOW, resources=[f"arn:{self.partition}:ecr:{self.region}:{self.account}:repository/{dep_mod}*"], ), iam.PolicyStatement( actions=[ "batch:UntagResource", "batch:DeregisterJobDefinition", "batch:TerminateJob", "batch:CancelJob", "batch:SubmitJob", "batch:RegisterJobDefinition", "batch:TagResource", ], effect=iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:batch:{self.region}:{self.account}:job-queue/addf*", f"arn:{self.partition}:batch:{self.region}:{self.account}:job-definition/*", f"arn:{self.partition}:batch:{self.region}:{self.account}:job/*", ], ), iam.PolicyStatement( actions=[ "iam:PassRole", ], effect=iam.Effect.ALLOW, resources=[ f"arn:{self.partition}:iam::{self.account}:role/addf*", ], ), iam.PolicyStatement( actions=[ "batch:Describe*", "batch:List*", ], effect=iam.Effect.ALLOW, resources=[ "*", ], ), ] dag_document = iam.PolicyDocument(statements=policy_statements) batch_role_name = f"{dep_mod}-dag-role" self.dag_role = iam.Role( self, f"dag-role-{dep_mod}", assumed_by=iam.CompositePrincipal( iam.ArnPrincipal(self.mwaa_exec_role), iam.ServicePrincipal("ecs-tasks.amazonaws.com"), ), inline_policies={"DagPolicyDocument": dag_document}, managed_policies=[ iam.ManagedPolicy.from_aws_managed_policy_name("service-role/AmazonECSTaskExecutionRolePolicy") ], role_name=batch_role_name, path="/", ) ec2Role = iam.Role( self, "BatchEC2Role", assumed_by=iam.CompositePrincipal(iam.ServicePrincipal("ec2.amazonaws.com")), managed_policies=[ iam.ManagedPolicy.from_aws_managed_policy_name("service-role/AmazonEC2ContainerServiceforEC2Role"), ], ) ec2IAMProfile = iam.CfnInstanceProfile(self, "BatchEC2RoleInstanceProfile", roles=[ec2Role.role_name]) batchSG = ec2.SecurityGroup( self, "BatchSG", vpc=self.vpc, allow_all_outbound=True, description="Batch SG", ) batchSG.add_egress_rule(ec2.Peer.ipv4(self.vpc.vpc_cidr_block), ec2.Port.all_tcp()) # Creates Compute Env conditionally batch_compute_config = batch_compute.get("batch_compute_config") on_demand_compute_env_list = [] spot_compute_env_list = [] fargate_compute_env_list = [] if batch_compute_config: for batchenv in batch_compute_config: if batchenv.get("compute_type").upper().startswith("ON"): instance_types_context = batchenv.get("instance_types") instance_types = [] if instance_types_context: for value in instance_types_context: instance_type = ec2.InstanceType(value) instance_types.append(instance_type) on_demand_compute_env = batch.ComputeEnvironment( self, f"{dep_mod}-OnDemandComputeEnv-{batchenv.get('env_name')}", compute_resources=batch.ComputeResources( instance_role=ec2IAMProfile.attr_arn, vpc=self.vpc, instance_types=instance_types if instance_types else None, maxv_cpus=batchenv.get("max_vcpus") if batchenv.get("max_vcpus") else "256", # type: ignore minv_cpus=0, type=batch.ComputeResourceType.ON_DEMAND, vpc_subnets=ec2.SubnetSelection(subnets=self.private_subnets), security_groups=[batchSG], ), ) on_demand_compute_env_list.append( batch.JobQueueComputeEnvironment( compute_environment=on_demand_compute_env, order=int(batchenv.get("order")), ) ) elif batchenv.get("compute_type").upper() == "SPOT": instance_types_context = batchenv.get("instance_types") instance_types = [] if instance_types_context: for value in instance_types_context: instance_type = ec2.InstanceType(value) instance_types.append(instance_type) spot_compute_env = batch.ComputeEnvironment( self, f"{dep_mod}-SpotComputeEnv-{batchenv.get('env_name')}", compute_resources=batch.ComputeResources( instance_role=ec2IAMProfile.attr_arn, vpc=self.vpc, instance_types=instance_types if instance_types else None, maxv_cpus=batchenv.get("max_vcpus") if batchenv.get("max_vcpus") else "256", # type: ignore minv_cpus=0, type=batch.ComputeResourceType.SPOT, vpc_subnets=ec2.SubnetSelection(subnets=self.private_subnets), security_groups=[batchSG], allocation_strategy=batch.AllocationStrategy("SPOT_CAPACITY_OPTIMIZED"), ), ) spot_compute_env_list.append( batch.JobQueueComputeEnvironment( compute_environment=spot_compute_env, order=int(batchenv.get("order")), ) ) else: fargate_compute_env = batch.ComputeEnvironment( self, f"{dep_mod}-FargateJobEnv-{batchenv.get('env_name')}", compute_resources=batch.ComputeResources( type=batch.ComputeResourceType.FARGATE, vpc_subnets=ec2.SubnetSelection(subnets=self.private_subnets), vpc=self.vpc, ), ) fargate_compute_env_list.append( batch.JobQueueComputeEnvironment( compute_environment=fargate_compute_env, order=int(batchenv.get("order")), ) ) if on_demand_compute_env_list: self.on_demand_jobqueue = batch.JobQueue( self, f"{dep_mod}-OndemandJobQueue", compute_environments=on_demand_compute_env_list, job_queue_name=f"{dep_mod}-OnDemandJobQueue", priority=1, ) if spot_compute_env_list: self.spot_jobqueue = batch.JobQueue( self, f"{dep_mod}-SpotJobQueue", compute_environments=spot_compute_env_list, job_queue_name=f"{dep_mod}-SpotJobQueue", priority=1, ) if fargate_compute_env_list: self.fargate_jobqueue = batch.JobQueue( self, f"{dep_mod}-FargateJobQueue", compute_environments=fargate_compute_env_list, job_queue_name=f"{dep_mod}-FargateJobQueue", priority=1, ) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks()) NagSuppressions.add_stack_suppressions( self, apply_to_nested_stacks=True, suppressions=[ { # type: ignore "id": "AwsSolutions-IAM4", "reason": "Managed Policies are for service account roles only", }, { # type: ignore "id": "AwsSolutions-IAM5", "reason": "Resource access restriced to ADDF resources", }, ], )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/app.py
modules/simulations/batch-managed/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import json import os from aws_cdk import App, CfnOutput, Environment from stack import BatchDags project_name = os.getenv("SEEDFARMER_PROJECT_NAME", "") deployment_name = os.getenv("SEEDFARMER_DEPLOYMENT_NAME", "") module_name = os.getenv("SEEDFARMER_MODULE_NAME", "") def _param(name: str) -> str: return f"SEEDFARMER_PARAMETER_{name}" vpc_id = os.getenv(_param("VPC_ID")) # required private_subnet_ids = json.loads(os.getenv(_param("PRIVATE_SUBNET_IDS"), "")) # required batch_compute = json.loads(os.getenv(_param("BATCH_COMPUTE"), "")) # required mwaa_exec_role = os.getenv(_param("MWAA_EXEC_ROLE")) if not vpc_id: raise Exception("missing input parameter vpc-id") if not private_subnet_ids: raise Exception("missing input parameter private-subnet-ids") if not batch_compute: raise ValueError("Batch Compute Configuration is missing.") if not mwaa_exec_role: raise ValueError("MWAA Execution Role is missing.") app = App() stack = BatchDags( scope=app, id=f"{project_name}-{deployment_name}-{module_name}", project_name=project_name, deployment_name=deployment_name, module_name=module_name, mwaa_exec_role=mwaa_exec_role, vpc_id=vpc_id, private_subnet_ids=private_subnet_ids, batch_compute=batch_compute, env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( scope=stack, id="metadata", value=stack.to_json_string( { "DagRoleArn": stack.dag_role.role_arn, "OnDemandJobQueueArn": stack.on_demand_jobqueue.job_queue_arn if hasattr(stack, "on_demand_jobqueue") else "QUEUE NOT CREATED", "SpotJobQueueArn": stack.spot_jobqueue.job_queue_arn if hasattr(stack, "spot_jobqueue") else "QUEUE NOT CREATED", "FargateJobQueueArn": stack.fargate_jobqueue.job_queue_arn if hasattr(stack, "fargate_jobqueue") else "QUEUE NOT CREATED", } ), ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/simulation_batch_dags/batch_simple_mock_fargate.py
modules/simulations/batch-managed/simulation_batch_dags/batch_simple_mock_fargate.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import json import logging import os import random import string import textwrap from datetime import timedelta from typing import TypeVar import boto3 from airflow import DAG, settings from airflow.contrib.hooks.aws_hook import AwsHook from airflow.exceptions import AirflowException from airflow.models import Connection from airflow.operators.python import PythonOperator from airflow.providers.amazon.aws.operators.batch import AwsBatchOperator from airflow.utils.dates import days_ago from boto3.session import Session from mypy_boto3_batch.client import BatchClient from simulation_batch_dags import batch_dag_config ValueType = TypeVar("ValueType") DAG_ID = os.path.basename(__file__).replace(".py", "") TASK_DEF_XCOM_KEY = "job_definition_name" DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("DEBUG") def try_create_aws_conn(**kwargs): conn_id = "aws_batch" try: AwsHook.get_connection(conn_id) except AirflowException: extra = json.dumps({"role_arn": batch_dag_config.DAG_ROLE}, indent=2) conn = Connection(conn_id=conn_id, conn_type="aws", host="", schema="", login="", extra=extra) try: session = settings.Session() session.add(conn) session.commit() finally: session.close() def get_job_queue_name() -> str: """get_job_queue_name retrieves the the available JobQueues created based on the inputs of the batch-compute manifest file. """ # ON_DEMAND_JOB_QUEUE_ARN = batch_dag_config.ON_DEMAND_JOB_QUEUE_ARN # consume if created # SPOT_JOB_QUEUE_ARN = batch_dag_config.SPOT_JOB_QUEUE_ARN # consume if created FARGATE_JOB_QUEUE_ARN = batch_dag_config.FARGATE_JOB_QUEUE_ARN # consume if created return FARGATE_JOB_QUEUE_ARN.split("/")[-1] def get_job_name() -> str: v = "".join(random.choice(string.ascii_lowercase) for i in range(6)) return f"addf-{batch_dag_config.DEPLOYMENT_NAME}-{batch_dag_config.MODULE_NAME}-simplemock-fargate-job-{v}" def get_job_def_name() -> str: # v = "".join(random.choice(string.ascii_lowercase) for i in range(6)) # return f"addf-{batch_dag_config.DEPLOYMENT_NAME}-{batch_dag_config.MODULE_NAME}-jobdef-{v}" return f"addf-{batch_dag_config.DEPLOYMENT_NAME}-{batch_dag_config.MODULE_NAME}-simplemock-fargate-jobdef" def get_batch_client() -> BatchClient: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=batch_dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session.client("batch") def register_job_definition_fargate(ti, job_def_name: str) -> str: client = get_batch_client() resp = client.register_job_definition( jobDefinitionName=job_def_name, type="container", containerProperties={ "image": "ubuntu", "jobRoleArn": batch_dag_config.DAG_ROLE, "executionRoleArn": batch_dag_config.DAG_ROLE, "environment": [ { "name": "AWS_DEFAULT_REGION", "value": batch_dag_config.REGION, } ], "resourceRequirements": [ {"value": ".25", "type": "VCPU"}, {"value": "1024", "type": "MEMORY"}, ], }, propagateTags=True, timeout={"attemptDurationSeconds": 60}, platformCapabilities=["FARGATE"], ) ti.xcom_push(key=TASK_DEF_XCOM_KEY, value=resp["jobDefinitionName"]) def deregister_job_definition(ti, job_def_name: str) -> None: client = get_batch_client() response = client.deregister_job_definition(jobDefinition=job_def_name) ti.xcom_push(key="TASK_DEF_DEREGISTER_XCOM_KEY", value=response["ResponseMetadata"]) with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: total_simulations = 50 parallelism = 10 job_definition_name = get_job_def_name() job_name = get_job_name() queue_name = get_job_queue_name() create_aws_conn = PythonOperator( task_id="try_create_aws_conn", python_callable=try_create_aws_conn, dag=dag, ) register_batch_job_defintion = PythonOperator( task_id="register_batch_job_definition", dag=dag, provide_context=True, python_callable=register_job_definition_fargate, op_kwargs={"job_def_name": job_definition_name}, ) submit_batch_job = AwsBatchOperator( task_id="submit_batch_job", job_name=job_name, job_queue=queue_name, aws_conn_id="aws_batch", # job_definition="{{ task_instance.xcom_pull(task_ids='register_batch_job_defintion', key='job_definition_name') }}", # noqa: E501 job_definition="addf-local-simulations-batch-managed-simplemock-fargate-jobdef", overrides={ "command": [ "bash", "-c", textwrap.dedent( """\ #/usr/bin/env bash echo "[$(date)] Starting $JOB_NAME" TIC_COUNT=$RANDOM let "TIC_COUNT %= $MAX_SECONDS" FAILURE_CHECK=$RANDOM let "FAILURE_CHECK %= $FAILURE_SEED" echo "[$(date)] Random Runtime: $TIC_COUNT" echo "[$(date)] Random Failure: $FAILURE_CHECK" CURRENT_COUNT=0 while true; do CURRENT_COUNT=$((CURRENT_COUNT + 1)) if [ "$CURRENT_COUNT" -ge "$TIC_COUNT" ]; then break fi sleep 1 done if [ "$FAILURE_CHECK" -eq "0" ]; then echo "[$(date)] Failure" && exit 1 else echo "[$(date)] Complete" && exit 0 fi """ ), ], "environment": [ {"name": "JOB_NAME", "value": "Simple Batch Mock"}, {"name": "MAX_SECONDS", "value": "120"}, {"name": "FAILURE_SEED", "value": "20"}, {"name": "DEBUG", "value": "true"}, {"name": "AWS_ACCOUNT_ID", "value": batch_dag_config.ACCOUNT_ID}, ], }, ) deregister_batch_job_defintion = PythonOperator( task_id="deregister_batch_job_defintion", dag=dag, provide_context=True, op_kwargs={ # "job_def_name": "{{ task_instance.xcom_pull(task_ids='deregister_batch_job_defintion', key='job_definition_name') }}" # noqa: E501 "job_def_name": "addf-local-simulations-batch-managed-simplemock-fargate-jobdef" }, python_callable=deregister_job_definition, ) (create_aws_conn >> register_batch_job_defintion >> submit_batch_job >> deregister_batch_job_defintion)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/simulation_batch_dags/batch_simple_mock_ec2.py
modules/simulations/batch-managed/simulation_batch_dags/batch_simple_mock_ec2.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import json import logging import os import random import string import textwrap from datetime import timedelta from typing import TypeVar import boto3 from airflow import DAG, settings from airflow.contrib.hooks.aws_hook import AwsHook from airflow.exceptions import AirflowException from airflow.models import Connection from airflow.operators.python import PythonOperator from airflow.providers.amazon.aws.operators.batch import AwsBatchOperator from airflow.utils.dates import days_ago from boto3.session import Session from mypy_boto3_batch.client import BatchClient from simulation_batch_dags import batch_dag_config ValueType = TypeVar("ValueType") DAG_ID = os.path.basename(__file__).replace(".py", "") TASK_DEF_XCOM_KEY = "job_definition_arn" DEFAULT_ARGS = { "owner": "airflow", "depends_on_past": False, "email": ["airflow@example.com"], } logger = logging.getLogger("airflow") logger.setLevel("DEBUG") def try_create_aws_conn(**kwargs): conn_id = "aws_batch" try: AwsHook.get_connection(conn_id) except AirflowException: extra = json.dumps({"role_arn": batch_dag_config.DAG_ROLE}, indent=2) conn = Connection(conn_id=conn_id, conn_type="aws", host="", schema="", login="", extra=extra) try: session = settings.Session() session.add(conn) session.commit() finally: session.close() def get_job_queue_name() -> str: """get_job_queue_name retrieves the the available JobQueues created based on the inputs of the batch-compute manifest file. """ ON_DEMAND_JOB_QUEUE_ARN = batch_dag_config.ON_DEMAND_JOB_QUEUE_ARN # consume if created # SPOT_JOB_QUEUE_ARN = batch_dag_config.SPOT_JOB_QUEUE_ARN # consume if created # FARGATE_JOB_QUEUE_ARN = batch_dag_config.FARGATE_JOB_QUEUE_ARN # consume if created return ON_DEMAND_JOB_QUEUE_ARN.split("/")[-1] def get_job_name() -> str: v = "".join(random.choice(string.ascii_lowercase) for i in range(6)) return "-".join( [ batch_dag_config.PROJECT_NAME, batch_dag_config.DEPLOYMENT_NAME, batch_dag_config.MODULE_NAME, "simplemock-job", v, ] ) def get_job_def_name() -> str: return "-".join( [ batch_dag_config.PROJECT_NAME, batch_dag_config.DEPLOYMENT_NAME, batch_dag_config.MODULE_NAME, "simplemock-jobdef", ] ) def get_batch_client() -> BatchClient: sts_client = boto3.client("sts") response = sts_client.assume_role( RoleArn=batch_dag_config.DAG_ROLE, RoleSessionName="AssumeRoleSession1", ) session = Session( aws_access_key_id=response["Credentials"]["AccessKeyId"], aws_secret_access_key=response["Credentials"]["SecretAccessKey"], aws_session_token=response["Credentials"]["SessionToken"], ) return session.client("batch") def register_job_definition_on_demand(ti, job_def_name: str) -> str: client = get_batch_client() resp = client.register_job_definition( jobDefinitionName=job_def_name, type="container", containerProperties={ "image": "ubuntu", "jobRoleArn": batch_dag_config.DAG_ROLE, "environment": [ { "name": "AWS_DEFAULT_REGION", "value": batch_dag_config.REGION, } ], "resourceRequirements": [ {"value": "1", "type": "VCPU"}, {"value": "512", "type": "MEMORY"}, ], }, propagateTags=True, timeout={"attemptDurationSeconds": 60}, platformCapabilities=["EC2"], ) ti.xcom_push(key=TASK_DEF_XCOM_KEY, value=resp["jobDefinitionArn"]) def deregister_job_definition(ti, job_def_arn: str) -> None: client = get_batch_client() response = client.deregister_job_definition(jobDefinition=job_def_arn) ti.xcom_push(key="TASK_DEF_DEREGISTER_XCOM_KEY", value=response["ResponseMetadata"]) with DAG( dag_id=DAG_ID, default_args=DEFAULT_ARGS, dagrun_timeout=timedelta(hours=2), start_date=days_ago(1), # type: ignore schedule_interval="@once", ) as dag: total_simulations = 50 parallelism = 10 job_definition_name = get_job_def_name() job_name = get_job_name() queue_name = get_job_queue_name() create_aws_conn = PythonOperator( task_id="try_create_aws_conn", python_callable=try_create_aws_conn, dag=dag, ) register_batch_job_defintion = PythonOperator( task_id="register-batch-job-definition", dag=dag, provide_context=True, python_callable=register_job_definition_on_demand, op_kwargs={"job_def_name": job_definition_name}, ) submit_batch_job = AwsBatchOperator( task_id="submit_batch_job", job_name=job_name, job_queue=queue_name, aws_conn_id="aws_batch", # job_definition="{{ task_instance.xcom_pull(task_ids='register_batch_job_defintion', key='job_definition_name') }}", # noqa: E501 job_definition=f"{batch_dag_config.PROJECT_NAME}-local-simulations-batch-managed-simplemock-jobdef", # TODO overrides={ "command": [ "bash", "-c", textwrap.dedent( """\ #/usr/bin/env bash echo "[$(date)] Starting $JOB_NAME" TIC_COUNT=$RANDOM let "TIC_COUNT %= $MAX_SECONDS" FAILURE_CHECK=$RANDOM let "FAILURE_CHECK %= $FAILURE_SEED" echo "[$(date)] Random Runtime: $TIC_COUNT" echo "[$(date)] Random Failure: $FAILURE_CHECK" CURRENT_COUNT=0 while true; do CURRENT_COUNT=$((CURRENT_COUNT + 1)) if [ "$CURRENT_COUNT" -ge "$TIC_COUNT" ]; then break fi sleep 1 done if [ "$FAILURE_CHECK" -eq "0" ]; then echo "[$(date)] Failure" && exit 1 else echo "[$(date)] Complete" && exit 0 fi """ ), ], "environment": [ {"name": "JOB_NAME", "value": "Simple Batch Mock"}, {"name": "MAX_SECONDS", "value": "120"}, {"name": "FAILURE_SEED", "value": "20"}, {"name": "DEBUG", "value": "true"}, {"name": "AWS_ACCOUNT_ID", "value": batch_dag_config.ACCOUNT_ID}, ], }, ) deregister_batch_job_defintion = PythonOperator( task_id="deregister_batch_job_defintion", dag=dag, provide_context=True, op_kwargs={ # "job_def_arn": "{{ task_instance.xcom_pull(task_ids='deregister_batch_job_defintion', key='job_definition_arn') }}" # noqa: E501 "job_def_arn": f"{batch_dag_config.PROJECT_NAME}-local-simulations-batch-managed-simplemock-jobdef" }, python_callable=deregister_job_definition, ) (create_aws_conn >> register_batch_job_defintion >> submit_batch_job >> deregister_batch_job_defintion)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/simulation_batch_dags/batch_dag_config.py
modules/simulations/batch-managed/simulation_batch_dags/batch_dag_config.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # This file is populated with configurations information when the Module is deployed # Configuration parameters are exported as module level constants # # Example: # SOME_PARAMETER = 'some value'
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/simulation_batch_dags/__init__.py
modules/simulations/batch-managed/simulation_batch_dags/__init__.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License.
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/tests/test_stack.py
modules/simulations/batch-managed/tests/test_stack.py
def test_placeholder() -> None: return None
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/tests/__init__.py
modules/simulations/batch-managed/tests/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/images/simulation-mock/src/setup.py
modules/simulations/batch-managed/images/simulation-mock/src/setup.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from setuptools import find_packages, setup setup( name="simulation-mock", version="0.1.0", author="AWS Professional Services", author_email="aws-proserve-opensource@amazon.com", project_urls={"Org Site": "https://aws.amazon.com/professional-services/"}, packages=find_packages(include=["simulation_mock", "simulation_mock.*"]), python_requires=">=3.7, <3.14", install_requires=["boto3~=1.21.19", "platonic-sqs==1.0.1"], include_package_data=True, )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/__init__.py
modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/__init__.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import logging import os DEBUG_LOGGING_FORMAT = "[%(asctime)s][%(filename)-13s:%(lineno)3d][%(levelname)s][%(threadName)s] %(message)s" def get_logger() -> logging.Logger: debug = os.environ.get("DEBUG", "False").lower() in [ "true", "yes", "1", ] level = logging.DEBUG if debug else logging.INFO logging.basicConfig(level=level, format=DEBUG_LOGGING_FORMAT) logger: logging.Logger = logging.getLogger(__name__) logger.setLevel(level) if debug: logging.getLogger("boto3").setLevel(logging.ERROR) logging.getLogger("botocore").setLevel(logging.ERROR) logging.getLogger("urllib3").setLevel(logging.ERROR) return logger
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/sqs_manager.py
modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/sqs_manager.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import argparse import os import sys import time from contextlib import contextmanager from datetime import timedelta from typing import Iterator from platonic.queue.errors import MessageReceiveTimeout from platonic.sqs.queue import SQSReceiver from platonic.sqs.queue.message import SQSMessage from platonic.sqs.queue.types import ValueType from platonic.timeout import ConstantTimeout from simulation_mock import get_logger MESSAGE_RECEIVE_TIMEOUT_SECONDS = 90 PROCESSING_TIMEOUT_MINUTES = 5 FILE_CHECK_SLEEP_TIME_SECONDS = 30 DATA_FILE_NAME = "message.json" LOGGER = get_logger() class ProcessingTimeoutException(Exception): pass class SQSHeartbeatReceiver(SQSReceiver[ValueType]): def heartbeat(self, message: SQSMessage[ValueType], seconds: int) -> None: """Extend the visibility timeout of Messages being processed Parameters ---------- message : SQSMessage[ValueType] The SQS Message to extend seconds : int Number of seconds to extend the timeout """ self.client.change_message_visibility( QueueUrl=self.url, ReceiptHandle=message.receipt_handle, VisibilityTimeout=seconds, ) @contextmanager def acknowledgement(self, message: SQSMessage[ValueType]) -> Iterator[SQSMessage[ValueType]]: """Override of ``acknowledgement`` that won't acknowledge (delete) them message if an Exception is thrown Parameters ---------- message : SQSMessage[ValueType] The SQS Message to conditionally acknowledge Yields ------ SQSMessage[ValueType] The SQS Message Raises ------ e Any exception that was thrown """ try: # noqa: WPS501 yield message except Exception as e: LOGGER.debug("NAK: %s - %s", message, e) raise e else: LOGGER.debug("ACK: %s", message) self.acknowledge(message) def main(url: str, dir: str, single_message: bool) -> int: LOGGER.info("QUEUE: %s", url) LOGGER.info("DIR: %s", dir) LOGGER.info("SINGLE_MESSAGE: %s", single_message) incoming_simulations = SQSHeartbeatReceiver[str]( url=url, timeout=ConstantTimeout(period=timedelta(seconds=MESSAGE_RECEIVE_TIMEOUT_SECONDS)), ) while True: try: with incoming_simulations.acknowledgement(incoming_simulations.receive()) as message: data_file = os.path.join(dir, "message.json") LOGGER.info("MSG: %s", message) LOGGER.info("DATA_FILE: %s", data_file) with open(data_file, "w") as file: file.write(message.value) timeout = ConstantTimeout(timedelta(minutes=PROCESSING_TIMEOUT_MINUTES)) with timeout.timer() as timer: while not timer.is_expired: if os.path.isfile(data_file): LOGGER.debug("EXISTS: %s", data_file) incoming_simulations.heartbeat( message=message, seconds=(MESSAGE_RECEIVE_TIMEOUT_SECONDS - 30), ) time.sleep(FILE_CHECK_SLEEP_TIME_SECONDS) else: LOGGER.debug("GONE: %s", data_file) break else: raise ProcessingTimeoutException("TIMEOUT") except MessageReceiveTimeout: LOGGER.info("EMPTY: %s", url) return 0 except ProcessingTimeoutException: LOGGER.error("TIMEOUT: %s", PROCESSING_TIMEOUT_MINUTES) if single_message: return 1 else: LOGGER.info("SUCCESS: %s", message) if single_message: return 0 if __name__ == "__main__": parser = argparse.ArgumentParser(description="Manage SQS Messages") parser.add_argument("--url", required=True, help="URL of the SQS Queue to manage") parser.add_argument("--dir", required=True, help="Directory to use/monitor for data files") parser.add_argument( "--single-message", action="store_true", help="Whether to retrieve/process a single message or process messages until the queue is empty", ) args = parser.parse_args() LOGGER.debug("ARGS: %s", args) sys.exit(main(url=args.url, dir=args.dir, single_message=args.single_message))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/simulator.py
modules/simulations/batch-managed/images/simulation-mock/src/simulation_mock/simulator.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"). # You may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import argparse import os import sys import time from datetime import timedelta from random import randint from platonic.timeout import ConstantTimeout from simulation_mock import get_logger FILE_CHECK_TIMEOUT_SECONDS = 120 DATA_FILE_NAME = "message.json" FILE_CHECK_SLEEP_TIME_SECONDS = 10 LOGGER = get_logger() class RandomFailureException(Exception): pass class FileCheckTimoutException(Exception): pass def main(dir: str, max_seconds: int, failure_seed: int) -> int: LOGGER.info("DIR: %s", dir) LOGGER.info("MAX_SECONDS: %s", max_seconds) LOGGER.info("FAILURE_SEED: %s", failure_seed) while True: try: timeout = ConstantTimeout(timedelta(seconds=FILE_CHECK_TIMEOUT_SECONDS)) with timeout.timer() as timer: data_file = os.path.join(dir, "message.json") LOGGER.info("DATA_FILE: %s", data_file) while not timer.is_expired: if os.path.isfile(data_file): LOGGER.info("EXISTS: %s", data_file) sleep_time = randint(0, max_seconds) LOGGER.info("RANDOM_SLEEP: %s", sleep_time) time.sleep(sleep_time) if (randint(0, 32768) % failure_seed) == 0: raise RandomFailureException("RANDOM") else: LOGGER.info("REMOVING: %s", data_file) os.remove(data_file) break else: LOGGER.debug("NOT_EXISTS: %s", data_file) time.sleep(FILE_CHECK_SLEEP_TIME_SECONDS) else: raise FileCheckTimoutException("TIMEOUT") except RandomFailureException: LOGGER.error("RANDOM_FAILURE") return 1 except FileCheckTimoutException: LOGGER.info("EXISTS_TIMEOUT: %s", data_file) return 0 if __name__ == "__main__": parser = argparse.ArgumentParser(description="Mock simulation") parser.add_argument("--dir", required=True, help="Directory to use/monitor for data files") parser.add_argument( "--max-seconds", required=True, type=int, help="Max runtime to execute mock simulation", ) parser.add_argument( "--failure-seed", required=True, type=int, help="Seed number to determine random failures", ) args = parser.parse_args() LOGGER.debug("ARGS: %s", args) sys.exit(main(dir=args.dir, max_seconds=args.max_seconds, failure_seed=args.failure_seed))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/stack.py
modules/service-catalog/stack.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from typing import Any, cast import aws_cdk.aws_iam as iam import aws_cdk.aws_s3_assets as assets import aws_cdk.aws_servicecatalog as servicecatalog import cdk_nag from aws_cdk import Aspects, Stack, Tags from constructs import Construct, IConstruct from service_catalog.products import products class ServiceCatalogStack(Stack): def __init__( self, scope: Construct, id: str, *, deployment_name: str, module_name: str, portfolio_access_role_arn: str, **kwargs: Any, ) -> None: super().__init__( scope, id, **kwargs, ) self.deployment_name = deployment_name self.module_name = module_name Tags.of(scope=cast(IConstruct, self)).add( key="Deployment", value=f"addf-{self.deployment_name}-{self.module_name}" ) self.portfolio = servicecatalog.Portfolio( self, "Portfolio", display_name="ADDF_Portfolio", provider_name="addf-admin", description="Portfolio for application templates provided by ADDF", message_language=servicecatalog.MessageLanguage.EN, ) account_root_principle = iam.Role( self, "AccountRootPrincipal", assumed_by=iam.AccountRootPrincipal(), ) self.portfolio.give_access_to_role(account_root_principle) if portfolio_access_role_arn is not None: self.portfolio_access_role = iam.Role.from_role_arn( self, "portfolio-access-role", portfolio_access_role_arn ) self.portfolio.give_access_to_role(self.portfolio_access_role) seed_code_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), "seed_code") for asset_dir_path in next(os.walk(seed_code_dir))[1]: asset_dir = os.path.basename(asset_dir_path) asset_name = "-" + asset_dir.replace("_", "-") app_asset = assets.Asset( self, f"AppAsset{asset_name}", path=os.path.join(seed_code_dir, asset_dir), ) product_name = "ApplicationTemplate" + asset_name product = servicecatalog.CloudFormationProduct( self, f"AppTemplateProductRef{asset_name}", product_name=product_name, owner="addf-admin", product_versions=[ servicecatalog.CloudFormationProductVersion( cloud_formation_template=servicecatalog.CloudFormationTemplate.from_product_stack( products[asset_dir]( self, f"AppTemplateProduct{app_asset}", code_asset=app_asset, ), ), ), ], ) Tags.of(product).add(key="sagemaker:studio-visibility", value="true") self.portfolio.add_product(product) Aspects.of(self).add(cdk_nag.AwsSolutionsChecks())
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/app.py
modules/service-catalog/app.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import os from typing import cast from aws_cdk import App, CfnOutput, Environment from stack import ServiceCatalogStack deployment_name = os.getenv("ADDF_DEPLOYMENT_NAME", "") module_name = os.getenv("ADDF_MODULE_NAME", "") app = App() def _param(name: str) -> str: return f"ADDF_PARAMETER_{name}" DEFAULT_PORTFOLIO_ACCESS_ROLE_ARN = None portfolio_access_role_arn = os.getenv(_param("PORTFOLIO_ACCESS_ROLE_ARN"), DEFAULT_PORTFOLIO_ACCESS_ROLE_ARN) stack = ServiceCatalogStack( scope=app, id=f"addf-{deployment_name}-{module_name}", deployment_name=deployment_name, module_name=module_name, portfolio_access_role_arn=cast(str, portfolio_access_role_arn), env=Environment( account=os.environ["CDK_DEFAULT_ACCOUNT"], region=os.environ["CDK_DEFAULT_REGION"], ), ) CfnOutput( stack, "PortfolioARN", value=stack.portfolio.portfolio_arn, ) if portfolio_access_role_arn is not None: CfnOutput( stack, "PortfolioAccessRoleName", value=stack.portfolio_access_role.role_name, ) CfnOutput( stack, "PortfolioAccessRoleARN", value=stack.portfolio_access_role.role_arn, ) app.synth(force=True)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/service_catalog/products.py
modules/service-catalog/service_catalog/products.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import aws_cdk.aws_codecommit as codecommit import aws_cdk.aws_s3_assets as assets import aws_cdk.aws_servicecatalog as servicecatalog from aws_cdk import CfnOutput, CfnParameter from constructs import Construct class MLopsAppTemplateProduct(servicecatalog.ProductStack): def __init__(self, scope: Construct, id: str, code_asset: assets.Asset) -> None: super().__init__(scope, id) sagemaker_project_name = CfnParameter( self, "SageMakerProjectName", type="String", description="Name of the project.", ) sagemaker_project_id = CfnParameter( self, "SageMakerProjectId", type="String", description="Service generated Id of the project.", ) prefix = f"{sagemaker_project_name.value_as_string}-{sagemaker_project_id.value_as_string}" CfnOutput( self, "AssetPath", value=code_asset.asset_path, ) CfnOutput( self, "AssetBucket", value=code_asset.bucket.bucket_arn, ) codecommit.Repository( self, id="Repo", repository_name=f"{prefix}-repository", code=codecommit.Code.from_asset(code_asset), ) # key should be the same as asset_dir, i.e. # mlops_app corresponds to modules/service-catalog/seed_code/mlops_app products = {"mlops_app": MLopsAppTemplateProduct}
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/service_catalog/__init__.py
modules/service-catalog/service_catalog/__init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/ml/stable_pipeline.py
modules/service-catalog/seed_code/mlops_app/ml/stable_pipeline.py
import json import os import boto3 import sagemaker import sagemaker.session from sagemaker.estimator import Estimator from sagemaker.inputs import TrainingInput from sagemaker.model import Model from sagemaker.model_metrics import MetricsSource, ModelMetrics from sagemaker.processing import ProcessingInput, ProcessingOutput, ScriptProcessor from sagemaker.sklearn.processing import SKLearnProcessor from sagemaker.workflow.condition_step import ConditionStep from sagemaker.workflow.conditions import ConditionLessThanOrEqualTo from sagemaker.workflow.functions import JsonGet from sagemaker.workflow.model_step import ModelStep from sagemaker.workflow.parameters import ParameterInteger, ParameterString from sagemaker.workflow.pipeline import Pipeline from sagemaker.workflow.pipeline_context import PipelineSession from sagemaker.workflow.properties import PropertyFile from sagemaker.workflow.steps import ProcessingStep, TrainingStep BASE_DIR = os.path.dirname(os.path.realpath(__file__)) def get_role(sagemaker_session): return os.getenv( "SAGEMAKER_PIPELINE_ROLE_ARN", sagemaker.session.get_execution_role(sagemaker_session), ) def get_sagemaker_client(region): """Gets the sagemaker client. Args: region: the aws region to start the session default_bucket: the bucket to use for storing the artifacts Returns: `sagemaker.session.Session instance """ boto_session = boto3.Session(region_name=region) sagemaker_client = boto_session.client("sagemaker") return sagemaker_client def get_session(region, default_bucket): """Gets the sagemaker session based on the region. Args: region: the aws region to start the session default_bucket: the bucket to use for storing the artifacts Returns: `sagemaker.session.Session instance """ boto_session = boto3.Session(region_name=region) sagemaker_client = boto_session.client("sagemaker") runtime_client = boto_session.client("sagemaker-runtime") return sagemaker.session.Session( boto_session=boto_session, sagemaker_client=sagemaker_client, sagemaker_runtime_client=runtime_client, default_bucket=default_bucket, ) def get_pipeline_session(region, default_bucket): """Gets the pipeline session based on the region. Args: region: the aws region to start the session default_bucket: the bucket to use for storing the artifacts Returns: PipelineSession instance """ boto_session = boto3.Session(region_name=region) sagemaker_client = boto_session.client("sagemaker") return PipelineSession( boto_session=boto_session, sagemaker_client=sagemaker_client, default_bucket=default_bucket, ) def get_pipeline_custom_tags(new_tags, region, sagemaker_project_arn=None): try: sm_client = get_sagemaker_client(region) response = sm_client.list_tags( ResourceArn=sagemaker_project_arn, ) project_tags = response["Tags"] for project_tag in project_tags: new_tags.append(project_tag) except Exception as e: print(f"Error getting project tags: {e}") return new_tags def get_pipeline( region, sagemaker_project_arn=None, role=None, default_bucket=None, use_case_id="abalone", model_package_group_name="abalone-stable-models", base_job_prefix="abalone", ): """Gets a SageMaker ML Pipeline instance working with with Blazeface. Args: region: AWS region to create and run the pipeline. role: IAM role to create and run steps and pipeline. default_bucket: the bucket to use for storing the artifacts Returns: an instance of a pipeline """ config_file = open("../.sagemaker-code-config") sagemaker_code_config = json.load(config_file) sagemaker_pipeline_name = sagemaker_code_config["sagemakerPipelineName"] pipeline_name = f"{sagemaker_pipeline_name}-stable" model_package_group_name = f"{pipeline_name}-models" sagemaker_session = get_session(region, default_bucket) if role is None: role = get_role(sagemaker_session) pipeline_session = get_pipeline_session(region, default_bucket) # parameters for pipeline execution processing_instance_count = ParameterInteger(name="ProcessingInstanceCount", default_value=1) model_approval_status = ParameterString( name="ModelApprovalStatus", default_value="PendingManualApproval", ) input_data = ParameterString( name="InputDataUrl", default_value=f"s3://sagemaker-servicecatalog-seedcode-{region}/dataset/abalone-dataset.csv", ) processing_instance_type = "ml.m5.xlarge" training_instance_type = "ml.m5.xlarge" # processing step for feature engineering sklearn_processor = SKLearnProcessor( framework_version="0.23-1", instance_type=processing_instance_type, instance_count=processing_instance_count, base_job_name=f"{base_job_prefix}/sklearn-abalone-preprocess", sagemaker_session=pipeline_session, role=role, ) step_args = sklearn_processor.run( outputs=[ ProcessingOutput(output_name="train", source="/opt/ml/processing/train"), ProcessingOutput(output_name="validation", source="/opt/ml/processing/validation"), ProcessingOutput(output_name="test", source="/opt/ml/processing/test"), ], code=os.path.join(BASE_DIR, "scripts/preprocess.py"), arguments=["--input-data", input_data], ) step_process = ProcessingStep( name="PreprocessAbaloneData", step_args=step_args, ) # training step for generating model artifacts model_path = f"s3://{sagemaker_session.default_bucket()}/{base_job_prefix}/AbaloneTrain" image_uri = sagemaker.image_uris.retrieve( framework="xgboost", region=region, version="1.0-1", py_version="py3", instance_type=training_instance_type, ) xgb_train = Estimator( image_uri=image_uri, instance_type=training_instance_type, instance_count=1, output_path=model_path, base_job_name=f"{base_job_prefix}/abalone-train", sagemaker_session=pipeline_session, role=role, ) xgb_train.set_hyperparameters( objective="reg:linear", num_round=50, max_depth=5, eta=0.2, gamma=4, min_child_weight=6, subsample=0.7, silent=0, ) step_args = xgb_train.fit( inputs={ "train": TrainingInput( s3_data=step_process.properties.ProcessingOutputConfig.Outputs["train"].S3Output.S3Uri, content_type="text/csv", ), "validation": TrainingInput( s3_data=step_process.properties.ProcessingOutputConfig.Outputs["validation"].S3Output.S3Uri, content_type="text/csv", ), }, ) step_train = TrainingStep( name="TrainAbaloneModel", step_args=step_args, ) # processing step for evaluation script_eval = ScriptProcessor( image_uri=image_uri, command=["python3"], instance_type=processing_instance_type, instance_count=1, base_job_name=f"{base_job_prefix}/script-abalone-eval", sagemaker_session=pipeline_session, role=role, ) step_args = script_eval.run( inputs=[ ProcessingInput( source=step_train.properties.ModelArtifacts.S3ModelArtifacts, destination="/opt/ml/processing/model", ), ProcessingInput( source=step_process.properties.ProcessingOutputConfig.Outputs["test"].S3Output.S3Uri, destination="/opt/ml/processing/test", ), ], outputs=[ ProcessingOutput(output_name="evaluation", source="/opt/ml/processing/evaluation"), ], code=os.path.join(BASE_DIR, "scripts/evaluate.py"), ) evaluation_report = PropertyFile( name="AbaloneEvaluationReport", output_name="evaluation", path="evaluation.json", ) step_eval = ProcessingStep( name="EvaluateAbaloneModel", step_args=step_args, property_files=[evaluation_report], ) # register model step that will be conditionally executed model_metrics = ModelMetrics( model_statistics=MetricsSource( s3_uri="{}/evaluation.json".format( step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]["S3Output"]["S3Uri"], ), content_type="application/json", ), ) model = Model( image_uri=image_uri, model_data=step_train.properties.ModelArtifacts.S3ModelArtifacts, sagemaker_session=pipeline_session, role=role, ) step_args = model.register( content_types=["text/csv"], response_types=["text/csv"], inference_instances=["ml.t2.medium", "ml.m5.large"], transform_instances=["ml.m5.large"], model_package_group_name=model_package_group_name, approval_status=model_approval_status, model_metrics=model_metrics, ) step_register = ModelStep( name="RegisterAbaloneModel", step_args=step_args, ) # condition step for evaluating model quality and branching execution cond_lte = ConditionLessThanOrEqualTo( left=JsonGet( step_name=step_eval.name, property_file=evaluation_report, json_path="regression_metrics.mse.value", ), right=6.0, ) step_cond = ConditionStep( name="CheckMSEAbaloneEvaluation", conditions=[cond_lte], if_steps=[step_register], else_steps=[], ) # pipeline instance pipeline = Pipeline( name=pipeline_name, parameters=[ processing_instance_type, processing_instance_count, training_instance_type, model_approval_status, input_data, ], steps=[step_process, step_train, step_eval, step_cond], sagemaker_session=pipeline_session, ) return pipeline, role def upsert(region="us-east-2"): pipeline, role = get_pipeline(region=region) pipeline.upsert( role_arn=role, description=role, ) if __name__ == "__main__": upsert()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/ml/__init__.py
modules/service-catalog/seed_code/mlops_app/ml/__init__.py
# __init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/ml/scripts/__init__.py
modules/service-catalog/seed_code/mlops_app/ml/scripts/__init__.py
# __init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/ml/scripts/evaluate.py
modules/service-catalog/seed_code/mlops_app/ml/scripts/evaluate.py
"""Evaluation script for measuring mean squared error.""" import json import logging import pathlib import pickle import tarfile import numpy as np import pandas as pd import xgboost from sklearn.metrics import mean_squared_error logger = logging.getLogger() logger.setLevel(logging.INFO) logger.addHandler(logging.StreamHandler()) if __name__ == "__main__": logger.debug("Starting evaluation.") model_path = "/opt/ml/processing/model/model.tar.gz" with tarfile.open(model_path) as tar: tar.extractall(path=".") logger.debug("Loading xgboost model.") model = pickle.load(open("xgboost-model", "rb")) logger.debug("Reading test data.") test_path = "/opt/ml/processing/test/test.csv" df = pd.read_csv(test_path, header=None) logger.debug("Reading test data.") y_test = df.iloc[:, 0].to_numpy() df.drop(df.columns[0], axis=1, inplace=True) X_test = xgboost.DMatrix(df.values) logger.info("Performing predictions against test data.") predictions = model.predict(X_test) logger.debug("Calculating mean squared error.") mse = mean_squared_error(y_test, predictions) std = np.std(y_test - predictions) report_dict = { "regression_metrics": { "mse": { "value": mse, "standard_deviation": std, }, }, } output_dir = "/opt/ml/processing/evaluation" pathlib.Path(output_dir).mkdir(parents=True, exist_ok=True) logger.info("Writing out evaluation report with mse: %f", mse) evaluation_path = f"{output_dir}/evaluation.json" with open(evaluation_path, "w") as f: f.write(json.dumps(report_dict))
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/ml/scripts/preprocess.py
modules/service-catalog/seed_code/mlops_app/ml/scripts/preprocess.py
"""Feature engineers the abalone dataset.""" import argparse import logging import os import pathlib import boto3 import numpy as np import pandas as pd from sklearn.compose import ColumnTransformer from sklearn.impute import SimpleImputer from sklearn.pipeline import Pipeline from sklearn.preprocessing import OneHotEncoder, StandardScaler logger = logging.getLogger() logger.setLevel(logging.INFO) logger.addHandler(logging.StreamHandler()) # Since we get a headerless CSV file we specify the column names here. feature_columns_names = [ "sex", "length", "diameter", "height", "whole_weight", "shucked_weight", "viscera_weight", "shell_weight", ] label_column = "rings" feature_columns_dtype = { "sex": str, "length": np.float64, "diameter": np.float64, "height": np.float64, "whole_weight": np.float64, "shucked_weight": np.float64, "viscera_weight": np.float64, "shell_weight": np.float64, } label_column_dtype = {"rings": np.float64} def merge_two_dicts(x, y): """Merges two dicts, returning a new copy.""" z = x.copy() z.update(y) return z if __name__ == "__main__": logger.debug("Starting preprocessing.") parser = argparse.ArgumentParser() parser.add_argument("--input-data", type=str, required=True) args = parser.parse_args() base_dir = "/opt/ml/processing" pathlib.Path(f"{base_dir}/data").mkdir(parents=True, exist_ok=True) input_data = args.input_data bucket = input_data.split("/")[2] key = "/".join(input_data.split("/")[3:]) logger.info("Downloading data from bucket: %s, key: %s", bucket, key) fn = f"{base_dir}/data/abalone-dataset.csv" s3 = boto3.resource("s3") s3.Bucket(bucket).download_file(key, fn) logger.debug("Reading downloaded data.") df = pd.read_csv( fn, header=None, names=feature_columns_names + [label_column], dtype=merge_two_dicts(feature_columns_dtype, label_column_dtype), ) os.unlink(fn) logger.debug("Defining transformers.") numeric_features = list(feature_columns_names) numeric_features.remove("sex") numeric_transformer = Pipeline( steps=[ ("imputer", SimpleImputer(strategy="median")), ("scaler", StandardScaler()), ], ) categorical_features = ["sex"] categorical_transformer = Pipeline( steps=[ ("imputer", SimpleImputer(strategy="constant", fill_value="missing")), ("onehot", OneHotEncoder(handle_unknown="ignore")), ], ) preprocess = ColumnTransformer( transformers=[ ("num", numeric_transformer, numeric_features), ("cat", categorical_transformer, categorical_features), ], ) logger.info("Applying transforms.") y = df.pop("rings") X_pre = preprocess.fit_transform(df) y_pre = y.to_numpy().reshape(len(y), 1) X = np.concatenate((y_pre, X_pre), axis=1) logger.info("Splitting %d rows of data into train, validation, test datasets.", len(X)) np.random.shuffle(X) train, validation, test = np.split(X, [int(0.7 * len(X)), int(0.85 * len(X))]) logger.info("Writing out datasets to %s.", base_dir) pd.DataFrame(train).to_csv(f"{base_dir}/train/train.csv", header=False, index=False) pd.DataFrame(validation).to_csv( f"{base_dir}/validation/validation.csv", header=False, index=False, ) pd.DataFrame(test).to_csv(f"{base_dir}/test/test.csv", header=False, index=False)
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/infra/pipeline.py
modules/service-catalog/seed_code/mlops_app/infra/pipeline.py
# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 from typing import Any, Dict, cast import yaml from aws_cdk import Aws, Stack, Stage, pipelines from aws_cdk import aws_codebuild as codebuild from aws_cdk import aws_codecommit as codecommit from aws_cdk import aws_codepipeline as codepipeline from aws_cdk import aws_iam as iam from aws_cdk import aws_s3 as s3 from constructs import Construct from notifications.notifications_stack import NotificationsStack class PipelineStack(Stack): def __init__( self, scope: Construct, construct_id: str, code_repository_name: str, sagemaker_project_name: str, sagemaker_project_id: str, model_package_group_name: str, project_short_name: str, env_name: str, **kwargs: Any, ) -> None: super().__init__(scope, construct_id, **kwargs) source = pipelines.CodePipelineSource.code_commit( repository=codecommit.Repository.from_repository_name( self, f"source-repo-{sagemaker_project_name}-{sagemaker_project_id}", repository_name=code_repository_name, ), branch="main", ) code_build_role = iam.Role( self, f"codebuild-{sagemaker_project_name}-{sagemaker_project_id}", role_name=f"CodeBuildRole-{sagemaker_project_name}", assumed_by=iam.ServicePrincipal("codebuild.amazonaws.com"), managed_policies=[ iam.ManagedPolicy.from_aws_managed_policy_name("PowerUserAccess"), iam.ManagedPolicy.from_aws_managed_policy_name("IAMFullAccess"), ], ) # TODO narrow down sm_role = iam.Role( self, f"sm-role-{sagemaker_project_name}-{sagemaker_project_id}", role_name=f"sm-role-{sagemaker_project_name}-{sagemaker_project_id}", assumed_by=iam.ServicePrincipal("sagemaker.amazonaws.com"), managed_policies=[ iam.ManagedPolicy.from_aws_managed_policy_name( "AmazonSageMakerFullAccess", ), iam.ManagedPolicy.from_aws_managed_policy_name("AmazonS3FullAccess"), ], ) artifact_bucket_arn = self.node.try_get_context("artifact_bucket_arn") # use default value for s3 bucket if not provided through 'cdk.json' file if not artifact_bucket_arn: artifact_bucket_arn = f"arn:{self.partition}:s3:::sagemaker-{Stack.of(self).region}-{Stack.of(self).account}" artifact_bucket = s3.Bucket.from_bucket_arn( self, "code-pipeline-artifacts-bucket", artifact_bucket_arn, ) codepipeline_props = codepipeline.Pipeline( self, "CodepipelineProperty", artifact_bucket=artifact_bucket, pipeline_name=f"{project_short_name}-pipeline-{env_name}", ) self.pipeline = pipelines.CodePipeline( self, f"{sagemaker_project_name}-{sagemaker_project_id}-pipeline", code_pipeline=codepipeline_props, publish_assets_in_parallel=False, self_mutation=True, synth=pipelines.CodeBuildStep( "Synth", input=source, build_environment=codebuild.BuildEnvironment( build_image=codebuild.LinuxBuildImage.STANDARD_5_0, privileged=False, ), commands=[ "cd infra", "pip install -r requirements.txt", "npm install -g aws-cdk", 'cdk synth --app "python app.py"', ], role=code_build_role, primary_output_directory="infra/cdk.out", ), ) notification_stage_construct = NotificationStage( self, f"{sagemaker_project_name}-{sagemaker_project_id}-notifications-stage", sagemaker_project_name, sagemaker_project_id, model_package_group_name, project_short_name=project_short_name, env_name=env_name, ) notification_stage = self.pipeline.add_stage(notification_stage_construct) sm_pipelines_buildspec = self.convert_yaml_to_json("../buildspec.yaml") notification_stage.add_post( pipelines.CodeBuildStep( "SageMakerPipeline.Upsert", input=source, commands=[], build_environment=codebuild.BuildEnvironment( build_image=codebuild.LinuxBuildImage.STANDARD_5_0, environment_variables={ "SAGEMAKER_PROJECT_NAME": codebuild.BuildEnvironmentVariable( value=sagemaker_project_name, ), "SAGEMAKER_PROJECT_ID": codebuild.BuildEnvironmentVariable( value=sagemaker_project_id, ), "SAGEMAKER_PIPELINE_ROLE_ARN": codebuild.BuildEnvironmentVariable( value=sm_role.role_arn, ), "AWS_REGION": codebuild.BuildEnvironmentVariable( value=Aws.REGION, ), }, ), partial_build_spec=codebuild.BuildSpec.from_object( sm_pipelines_buildspec, ), role=code_build_role, ), ) artifact_bucket.grant_read_write(code_build_role) def convert_yaml_to_json(self, file_name: str) -> Dict[Any, Any]: with open(file_name, "r") as buildspec_yaml: return cast(Dict[Any, Any], yaml.safe_load(buildspec_yaml)) class NotificationStage(Stage): def __init__( self, scope: Construct, construct_id: str, sagemaker_project_name: str, sagemaker_project_id: str, model_package_group_name: str, project_short_name: str, env_name: str, **kwargs: Any, ) -> None: super().__init__(scope, construct_id, **kwargs) self.notification_stack = NotificationsStack( self, f"{sagemaker_project_name}-{sagemaker_project_id}-notif-stack", sagemaker_project_name=sagemaker_project_name, sagemaker_project_id=sagemaker_project_id, model_package_group_name=model_package_group_name, project_short_name=project_short_name, env_name=env_name, )
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/infra/__init__.py
modules/service-catalog/seed_code/mlops_app/infra/__init__.py
# __init__.py
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false
awslabs/autonomous-driving-data-framework
https://github.com/awslabs/autonomous-driving-data-framework/blob/deadc6a9cc30df557f10c750d7d4a0a751d42f53/modules/service-catalog/seed_code/mlops_app/infra/app.py
modules/service-catalog/seed_code/mlops_app/infra/app.py
#!/usr/bin/env python3 # Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. # SPDX-License-Identifier: Apache-2.0 import json import os from typing import cast import aws_cdk as cdk import boto3 from pipeline import PipelineStack config_file = open("../.sagemaker-code-config") sagemaker_code_config = json.load(config_file) sagemaker_project_name = sagemaker_code_config["sagemakerProjectName"] sagemaker_project_id = sagemaker_code_config["sagemakerProjectId"] sagemaker_repository_name = sagemaker_code_config["codeRepositoryName"] sagemaker_pipeline_name = sagemaker_code_config["sagemakerPipelineName"] project_short_name = sagemaker_code_config["projectShortName"] env_name = "dev" model_package_group_name = f"{sagemaker_pipeline_name}-{env_name}-models" def get_account() -> str: if "CDK_DEFAULT_ACCOUNT" in os.environ: return os.environ["CDK_DEFAULT_ACCOUNT"] return cast(str, boto3.client(service_name="sts").get_caller_identity().get("Account")) def get_region() -> str: if "CDK_DEFAULT_REGION" in os.environ: return os.environ["CDK_DEFAULT_REGION"] session = boto3.Session() if session.region_name is None: raise ValueError( "It is not possible to infer AWS REGION from your environment. Please pass the --region argument.", ) return str(session.region_name) app = cdk.App() PipelineStack( app, f"mlops-pipeline-{sagemaker_project_name}-{sagemaker_project_id}", code_repository_name=sagemaker_repository_name, sagemaker_project_name=sagemaker_project_name, sagemaker_project_id=sagemaker_project_id, model_package_group_name=model_package_group_name, project_short_name=project_short_name, env_name=env_name, env=cdk.Environment( account=get_account(), region=get_region(), ), ) app.synth()
python
Apache-2.0
deadc6a9cc30df557f10c750d7d4a0a751d42f53
2026-01-05T07:12:28.137953Z
false