Source code for airflow.providers.amazon.aws.transfers.redshift_to_s3

#
# 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.
"""Transfers data from AWS Redshift into a S3 Bucket."""
from typing import TYPE_CHECKING, Iterable, List, Mapping, Optional, Sequence, Union

from airflow.models import BaseOperator
from airflow.providers.amazon.aws.hooks.redshift_sql import RedshiftSQLHook
from airflow.providers.amazon.aws.hooks.s3 import S3Hook
from airflow.providers.amazon.aws.utils.redshift import build_credentials_block

if TYPE_CHECKING:
    from airflow.utils.context import Context


[docs]class RedshiftToS3Operator(BaseOperator): """ Executes an UNLOAD command to s3 as a CSV with headers .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:RedshiftToS3Operator` :param s3_bucket: reference to a specific S3 bucket :param s3_key: reference to a specific S3 key. If ``table_as_file_name`` is set to False, this param must include the desired file name :param schema: reference to a specific schema in redshift database Applicable when ``table`` param provided. :param table: reference to a specific table in redshift database Used when ``select_query`` param not provided. :param select_query: custom select query to fetch data from redshift database :param redshift_conn_id: reference to a specific redshift database :param aws_conn_id: reference to a specific S3 connection If the AWS connection contains 'aws_iam_role' in ``extras`` the operator will use AWS STS credentials with a token https://docs.aws.amazon.com/redshift/latest/dg/copy-parameters-authorization.html#copy-credentials :param verify: Whether or not to verify SSL certificates for S3 connection. By default SSL certificates are verified. You can provide the following values: - ``False``: do not validate SSL certificates. SSL will still be used (unless use_ssl is False), but SSL certificates will not be verified. - ``path/to/cert/bundle.pem``: A filename of the CA cert bundle to uses. You can specify this argument if you want to use a different CA cert bundle than the one used by botocore. :param unload_options: reference to a list of UNLOAD options :param autocommit: If set to True it will automatically commit the UNLOAD statement. Otherwise it will be committed right before the redshift connection gets closed. :param include_header: If set to True the s3 file contains the header columns. :param parameters: (optional) the parameters to render the SQL query with. :param table_as_file_name: If set to True, the s3 file will be named as the table. Applicable when ``table`` param provided. """
[docs] template_fields: Sequence[str] = ( 's3_bucket', 's3_key', 'schema', 'table', 'unload_options', 'select_query',
)
[docs] template_ext: Sequence[str] = ('.sql',)
[docs] template_fields_renderers = {'select_query': 'sql'}
[docs] ui_color = '#ededed'
def __init__( self, *, s3_bucket: str, s3_key: str, schema: Optional[str] = None, table: Optional[str] = None, select_query: Optional[str] = None, redshift_conn_id: str = 'redshift_default', aws_conn_id: str = 'aws_default', verify: Optional[Union[bool, str]] = None, unload_options: Optional[List] = None, autocommit: bool = False, include_header: bool = False, parameters: Optional[Union[Mapping, Iterable]] = None, table_as_file_name: bool = True, # Set to True by default for not breaking current workflows **kwargs, ) -> None: super().__init__(**kwargs) self.s3_bucket = s3_bucket self.s3_key = f'{s3_key}/{table}_' if (table and table_as_file_name) else s3_key self.schema = schema self.table = table self.redshift_conn_id = redshift_conn_id self.aws_conn_id = aws_conn_id self.verify = verify self.unload_options = unload_options or [] # type: List self.autocommit = autocommit self.include_header = include_header self.parameters = parameters self.table_as_file_name = table_as_file_name if select_query: self.select_query = select_query elif self.schema and self.table: self.select_query = f"SELECT * FROM {self.schema}.{self.table}" else: raise ValueError( 'Please provide both `schema` and `table` params or `select_query` to fetch the data.' ) if self.include_header and 'HEADER' not in [uo.upper().strip() for uo in self.unload_options]: self.unload_options = list(self.unload_options) + [ 'HEADER', ] def _build_unload_query( self, credentials_block: str, select_query: str, s3_key: str, unload_options: str ) -> str: return f""" UNLOAD ('{select_query}') TO 's3://{self.s3_bucket}/{s3_key}' credentials '{credentials_block}' {unload_options}; """
[docs] def execute(self, context: 'Context') -> None: redshift_hook = RedshiftSQLHook(redshift_conn_id=self.redshift_conn_id) conn = S3Hook.get_connection(conn_id=self.aws_conn_id) credentials_block = None if conn.extra_dejson.get('role_arn', False): credentials_block = f"aws_iam_role={conn.extra_dejson['role_arn']}" else: s3_hook = S3Hook(aws_conn_id=self.aws_conn_id, verify=self.verify) credentials = s3_hook.get_credentials() credentials_block = build_credentials_block(credentials) unload_options = '\n\t\t\t'.join(self.unload_options) unload_query = self._build_unload_query( credentials_block, self.select_query, self.s3_key, unload_options ) self.log.info('Executing UNLOAD command...') redshift_hook.run(unload_query, self.autocommit, parameters=self.parameters) self.log.info("UNLOAD command complete...")

Was this entry helpful?