dondaum commented on code in PR #39654: URL: https://github.com/apache/airflow/pull/39654#discussion_r1604601423
########## airflow/providers/amazon/aws/transfers/s3_to_dynamodb.py: ########## @@ -0,0 +1,256 @@ +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you 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 __future__ import annotations + +from typing import TYPE_CHECKING, Any, Literal, Sequence, TypedDict + +from botocore.exceptions import ClientError, WaiterError + +from airflow.exceptions import AirflowException +from airflow.models import BaseOperator +from airflow.providers.amazon.aws.hooks.dynamodb import DynamoDBHook + +if TYPE_CHECKING: + from airflow.utils.context import Context + + +class AttributeDefinition(TypedDict): + """Attribute Definition Type.""" + + AttributeName: str + AttributeType: Literal["S", "N", "B"] + + +class KeySchema(TypedDict): + """Key Schema Type.""" + + AttributeName: str + KeyType: Literal["HASH", "RANGE"] + + +class S3ToDynamoDBOperator(BaseOperator): + """Load Data from S3 into a DynamoDB. + + Data stored in S3 can be uploaded to a new or existing DynamoDB. Supported file formats CSV, DynamoDB JSON and + Amazon ION. + + + :param s3_bucket: The S3 bucket that is imported + :param s3_key: key prefix that imports single or multiple objects from S3 + :param dynamodb_table_name: Name of the table that shall be created + :param dynamodb_key_schema: Primary key and sort key. Each element represents one primary key + attribute. AttributeName is the name of the attribute. KeyType is the role for the attribute. Valid values + HASH or RANGE + :param dynamodb_attributes: Name of the attributes of a table. AttributeName is the name for the attribute + AttributeType is the data type for the attribute. Valid values for AttributeType are + S - attribute is of type String + N - attribute is of type Number + B - attribute is of type Binary + :param dynamodb_tmp_table_prefix: Prefix for the temporary DynamoDB table + :param delete_dynamodb_tmp_table: If set, deletes the temporary DynamoDB table that is used for staging + :param use_existing_table: Whether to import to an existing non new DynamoDB table. If set to + true data is loaded first into a temporary DynamoDB table, then retrieved as chunks into memory and loaded + into the target table + :param input_format: The format for the imported data. Valid values for InputFormat are CSV, DYNAMODB_JSON + or ION + :param billing_mode: Billing mode for the table. Valid values are PROVISIONED or PAY_PER_REQUEST + :param on_demand_throughput: Extra options for maximum number of read and write units + :param import_table_kwargs: Any additional optional import table parameters to pass, such as ClientToken, + InputCompressionType, or InputFormatOptions. See: + https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/dynamodb/client/import_table.html + :param import_table_creation_kwargs: Any additional optional import table creation parameters to pass, such as + ProvisionedThroughput, SSESpecification, or GlobalSecondaryIndexes. See: + https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/dynamodb/client/import_table.html + :param wait_for_completion: Whether to wait for cluster to stop + :param check_interval: Time in seconds to wait between status checks + :param max_attempts: Maximum number of attempts to check for job completion + :param aws_conn_id: The reference to the AWS connection details + """ + + template_fields: Sequence[str] = ( + "s3_bucket", + "s3_key", + "dynamodb_table_name", + "dynamodb_key_schema", + "dynamodb_attributes", + "dynamodb_tmp_table_prefix", + "delete_dynamodb_tmp_table", + "use_existing_table", + "input_format", + "billing_mode", + "import_table_kwargs", + "import_table_creation_kwargs", + ) + ui_color = "#e2e8f0" + + def __init__( + self, + *, + s3_bucket: str, + s3_key: str, + dynamodb_table_name: str, + dynamodb_key_schema: list[KeySchema], + dynamodb_attributes: list[AttributeDefinition] | None = None, + dynamodb_tmp_table_prefix: str = "tmp", + delete_dynamodb_tmp_table: bool = True, + use_existing_table: bool = False, + input_format: Literal["CSV", "DYNAMODB_JSON", "ION"] = "DYNAMODB_JSON", + billing_mode: Literal["PROVISIONED", "PAY_PER_REQUEST"] = "PAY_PER_REQUEST", + import_table_kwargs: dict[str, Any] | None = None, + import_table_creation_kwargs: dict[str, Any] | None = None, + wait_for_completion: bool = True, + check_interval: int = 30, + max_attempts: int = 240, + aws_conn_id: str | None = "aws_default", + **kwargs, + ) -> None: + super().__init__(**kwargs) + self.s3_bucket = s3_bucket + self.s3_key = s3_key + self.dynamodb_table_name = dynamodb_table_name + self.dynamodb_attributes = dynamodb_attributes + self.dynamodb_tmp_table_prefix = dynamodb_tmp_table_prefix + self.delete_dynamodb_tmp_table = delete_dynamodb_tmp_table + self.use_existing_table = use_existing_table + self.dynamodb_key_schema = dynamodb_key_schema + self.input_format = input_format + self.billing_mode = billing_mode + self.import_table_kwargs = import_table_kwargs + self.import_table_creation_kwargs = import_table_creation_kwargs + self.wait_for_completion = wait_for_completion + self.check_interval = check_interval + self.max_attempts = max_attempts + self.aws_conn_id = aws_conn_id + + @property + def tmp_table_name(self): + """Temporary table name.""" + return f"{self.dynamodb_tmp_table_prefix}_{self.dynamodb_table_name}" + + def _load_into_new_table(self, context: Context, tmp_table: bool = False) -> str: + """ + Import S3 key or keys into a new DynamoDB table. + + :param context: the current context of the task instance + :param tmp_table: creates DynamoDB Table with tmp prefix + :return: the Amazon resource number (ARN) + """ + dynamodb_hook = DynamoDBHook(aws_conn_id=self.aws_conn_id) + client = dynamodb_hook.client + table_name = self.tmp_table_name if tmp_table else self.dynamodb_table_name + + import_table_config = self.import_table_kwargs or {} + import_table_creation_config = self.import_table_creation_kwargs or {} + + try: + response = client.import_table( + S3BucketSource={ + "S3Bucket": self.s3_bucket, + "S3KeyPrefix": self.s3_key, + }, + InputFormat=self.input_format, + TableCreationParameters={ + "TableName": table_name, + "AttributeDefinitions": self.dynamodb_attributes, + "KeySchema": self.dynamodb_key_schema, + "BillingMode": self.billing_mode, + **import_table_creation_config, + }, + **import_table_config, + ) + except ClientError as e: + self.log.error("Error: failed to load from S3 into DynamoDB table. Error: %s", str(e)) + raise AirflowException(f"S3 load into DynamoDB table failed with error: {e}") + + if response["ImportTableDescription"]["ImportStatus"] == "FAILED": + raise AirflowException( + "S3 into Dynamodb job creation failed. Code: " + f"{response['ImportTableDescription']['FailureCode']}. " + f"Failure: {response['ImportTableDescription']['FailureMessage']}" + ) + + if self.wait_for_completion: + self.log.info("Waiting for S3 into Dynamodb job to complete") + waiter = dynamodb_hook.get_waiter("import_table") + try: + waiter.wait( + ImportArn=response["ImportTableDescription"]["ImportArn"], + WaiterConfig={"Delay": self.check_interval, "MaxAttempts": self.max_attempts}, + ) + except WaiterError: + status, error_code, error_msg = dynamodb_hook.get_import_status( + response["ImportTableDescription"]["ImportArn"] + ) + client.delete_table(TableName=table_name) + raise AirflowException( + f"S3 import into Dynamodb job failed: Status: {status}. Error: {error_code}. Error message: {error_msg}" + ) Review Comment: Got it. Thanks for the explanation. In fact, I implemented it so that in case of problems running the import (network or server side related), users might want to add a simple task retry. If we don't cleanup, retrying will result in another new error, because the table already exist. If that makes sense, I would change it to an optional flag `delete_on_error` with the default value set to False. I also added a unittest for the behavior of `delete_on_error`. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@airflow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
