airflow emr 源码
airflow emr 代码
文件路径：/airflow/providers/amazon/aws/hooks/emr.py
#
# 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
import json
import warnings
from time import sleep
from typing import Any, Callable
from botocore.exceptions import ClientError
from airflow.compat.functools import cached_property
from airflow.exceptions import AirflowException, AirflowNotFoundException
from airflow.providers.amazon.aws.hooks.base_aws import AwsBaseHook
class EmrHook(AwsBaseHook):
"""
Interact with Amazon Elastic MapReduce Service.
:param emr_conn_id: :ref:`Amazon Elastic MapReduce Connection <howto/connection:emr>`.
This attribute is only necessary when using
the :meth:`~airflow.providers.amazon.aws.hooks.emr.EmrHook.create_job_flow` method.
Additional arguments (such as ``aws_conn_id``) may be specified and
are passed down to the underlying AwsBaseHook.
.. seealso::
:class:`~airflow.providers.amazon.aws.hooks.base_aws.AwsBaseHook`
"""
conn_name_attr = 'emr_conn_id'
default_conn_name = 'emr_default'
conn_type = 'emr'
hook_name = 'Amazon Elastic MapReduce'
def __init__(self, emr_conn_id: str | None = default_conn_name, *args, **kwargs) -> None:
self.emr_conn_id = emr_conn_id
kwargs["client_type"] = "emr"
super().__init__(*args, **kwargs)
def get_cluster_id_by_name(self, emr_cluster_name: str, cluster_states: list[str]) -> str | None:
"""
Fetch id of EMR cluster with given name and (optional) states.
Will return only if single id is found.
:param emr_cluster_name: Name of a cluster to find
:param cluster_states: State(s) of cluster to find
:return: id of the EMR cluster
"""
response = self.get_conn().list_clusters(ClusterStates=cluster_states)
matching_clusters = list(
filter(lambda cluster: cluster['Name'] == emr_cluster_name, response['Clusters'])
)
if len(matching_clusters) == 1:
cluster_id = matching_clusters[0]['Id']
self.log.info('Found cluster name = %s id = %s', emr_cluster_name, cluster_id)
return cluster_id
elif len(matching_clusters) > 1:
raise AirflowException(f'More than one cluster found for name {emr_cluster_name}')
else:
self.log.info('No cluster found for name %s', emr_cluster_name)
return None
def create_job_flow(self, job_flow_overrides: dict[str, Any]) -> dict[str, Any]:
"""
Create and start running a new cluster (job flow).
This method uses ``EmrHook.emr_conn_id`` to receive the initial Amazon EMR cluster configuration.
If ``EmrHook.emr_conn_id`` is empty or the connection does not exist, then an empty initial
configuration is used.
:param job_flow_overrides: Is used to overwrite the parameters in the initial Amazon EMR configuration
cluster. The resulting configuration will be used in the boto3 emr client run_job_flow method.
.. seealso::
- :ref:`Amazon Elastic MapReduce Connection <howto/connection:emr>`
- `API RunJobFlow <https://docs.aws.amazon.com/emr/latest/APIReference/API_RunJobFlow.html>`_
- `boto3 emr client run_job_flow method <https://boto3.amazonaws.com/v1/documentation/\
api/latest/reference/services/emr.html#EMR.Client.run_job_flow>`_.
"""
config = {}
if self.emr_conn_id:
try:
emr_conn = self.get_connection(self.emr_conn_id)
except AirflowNotFoundException:
warnings.warn(
f"Unable to find {self.hook_name} Connection ID {self.emr_conn_id!r}, "
"using an empty initial configuration. If you want to get rid of this warning "
"message please provide a valid `emr_conn_id` or set it to None.",
UserWarning,
stacklevel=2,
)
else:
if emr_conn.conn_type and emr_conn.conn_type != self.conn_type:
warnings.warn(
f"{self.hook_name} Connection expected connection type {self.conn_type!r}, "
f"Connection {self.emr_conn_id!r} has conn_type={emr_conn.conn_type!r}. "
f"This connection might not work correctly.",
UserWarning,
stacklevel=2,
)
config = emr_conn.extra_dejson.copy()
config.update(job_flow_overrides)
response = self.get_conn().run_job_flow(**config)
return response
def test_connection(self):
"""
Return failed state for test Amazon Elastic MapReduce Connection (untestable).
We need to overwrite this method because this hook is based on
:class:`~airflow.providers.amazon.aws.hooks.base_aws.AwsGenericHook`,
otherwise it will try to test connection to AWS STS by using the default boto3 credential strategy.
"""
msg = (
f"{self.hook_name!r} Airflow Connection cannot be tested, by design it stores "
f"only key/value pairs and does not make a connection to an external resource."
)
return False, msg
@staticmethod
def get_ui_field_behaviour() -> dict[str, Any]:
"""Returns custom UI field behaviour for Amazon Elastic MapReduce Connection."""
return {
"hidden_fields": ["host", "schema", "port", "login", "password"],
"relabeling": {
"extra": "Run Job Flow Configuration",
},
"placeholders": {
"extra": json.dumps(
{
"Name": "MyClusterName",
"ReleaseLabel": "emr-5.36.0",
"Applications": [{"Name": "Spark"}],
"Instances": {
"InstanceGroups": [
{
"Name": "Primary node",
"Market": "SPOT",
"InstanceRole": "MASTER",
"InstanceType": "m5.large",
"InstanceCount": 1,
},
],
"KeepJobFlowAliveWhenNoSteps": False,
"TerminationProtected": False,
},
"StepConcurrencyLevel": 2,
},
indent=2,
),
},
}
class EmrServerlessHook(AwsBaseHook):
"""
Interact with EMR Serverless API.
Additional arguments (such as ``aws_conn_id``) may be specified and
are passed down to the underlying AwsBaseHook.
.. seealso::
:class:`~airflow.providers.amazon.aws.hooks.base_aws.AwsBaseHook`
"""
JOB_INTERMEDIATE_STATES = {'PENDING', 'RUNNING', 'SCHEDULED', 'SUBMITTED'}
JOB_FAILURE_STATES = {'FAILED', 'CANCELLING', 'CANCELLED'}
JOB_SUCCESS_STATES = {'SUCCESS'}
JOB_TERMINAL_STATES = JOB_SUCCESS_STATES.union(JOB_FAILURE_STATES)
APPLICATION_INTERMEDIATE_STATES = {'CREATING', 'STARTING', 'STOPPING'}
APPLICATION_FAILURE_STATES = {'STOPPED', 'TERMINATED'}
APPLICATION_SUCCESS_STATES = {'CREATED', 'STARTED'}
def __init__(self, *args: Any, **kwargs: Any) -> None:
kwargs["client_type"] = "emr-serverless"
super().__init__(*args, **kwargs)
@cached_property
def conn(self):
"""Get the underlying boto3 EmrServerlessAPIService client (cached)"""
return super().conn
# This method should be replaced with boto waiters which would implement timeouts and backoff nicely.
def waiter(
self,
get_state_callable: Callable,
get_state_args: dict,
parse_response: list,
desired_state: set,
failure_states: set,
object_type: str,
action: str,
countdown: int = 25 * 60,
check_interval_seconds: int = 60,
) -> None:
"""
Will run the sensor until it turns True.
:param get_state_callable: A callable to run until it returns True
:param get_state_args: Arguments to pass to get_state_callable
:param parse_response: Dictionary keys to extract state from response of get_state_callable
:param desired_state: Wait until the getter returns this value
:param failure_states: A set of states which indicate failure and should throw an
exception if any are reached before the desired_state
:param object_type: Used for the reporting string. What are you waiting for? (application, job, etc)
:param action: Used for the reporting string. What action are you waiting for? (created, deleted, etc)
:param countdown: Total amount of time the waiter should wait for the desired state
before timing out (in seconds). Defaults to 25 * 60 seconds.
:param check_interval_seconds: Number of seconds waiter should wait before attempting
to retry get_state_callable. Defaults to 60 seconds.
"""
response = get_state_callable(**get_state_args)
state: str = self.get_state(response, parse_response)
while state not in desired_state:
if state in failure_states:
raise AirflowException(f'{object_type.title()} reached failure state {state}.')
if countdown >= check_interval_seconds:
countdown -= check_interval_seconds
self.log.info('Waiting for %s to be %s.', object_type.lower(), action.lower())
sleep(check_interval_seconds)
state = self.get_state(get_state_callable(**get_state_args), parse_response)
else:
message = f'{object_type.title()} still not {action.lower()} after the allocated time limit.'
self.log.error(message)
raise RuntimeError(message)
def get_state(self, response, keys) -> str:
value = response
for key in keys:
if value is not None:
value = value.get(key, None)
return value
class EmrContainerHook(AwsBaseHook):
"""
Interact with AWS EMR Virtual Cluster to run, poll jobs and return job status
Additional arguments (such as ``aws_conn_id``) may be specified and
are passed down to the underlying AwsBaseHook.
.. seealso::
:class:`~airflow.providers.amazon.aws.hooks.base_aws.AwsBaseHook`
:param virtual_cluster_id: Cluster ID of the EMR on EKS virtual cluster
"""
INTERMEDIATE_STATES = (
"PENDING",
"SUBMITTED",
"RUNNING",
)
FAILURE_STATES = (
"FAILED",
"CANCELLED",
"CANCEL_PENDING",
)
SUCCESS_STATES = ("COMPLETED",)
TERMINAL_STATES = (
"COMPLETED",
"FAILED",
"CANCELLED",
"CANCEL_PENDING",
)
def __init__(self, *args: Any, virtual_cluster_id: str | None = None, **kwargs: Any) -> None:
super().__init__(client_type="emr-containers", *args, **kwargs) # type: ignore
self.virtual_cluster_id = virtual_cluster_id
def create_emr_on_eks_cluster(
self,
virtual_cluster_name: str,
eks_cluster_name: str,
eks_namespace: str,
tags: dict | None = None,
) -> str:
response = self.conn.create_virtual_cluster(
name=virtual_cluster_name,
containerProvider={
"id": eks_cluster_name,
"type": "EKS",
"info": {"eksInfo": {"namespace": eks_namespace}},
},
tags=tags or {},
)
if response['ResponseMetadata']['HTTPStatusCode'] != 200:
raise AirflowException(f'Create EMR EKS Cluster failed: {response}')
else:
self.log.info(
"Create EMR EKS Cluster success - virtual cluster id %s",
response['id'],
)
return response['id']
def submit_job(
self,
name: str,
execution_role_arn: str,
release_label: str,
job_driver: dict,
configuration_overrides: dict | None = None,
client_request_token: str | None = None,
tags: dict | None = None,
) -> str:
"""
Submit a job to the EMR Containers API and return the job ID.
A job run is a unit of work, such as a Spark jar, PySpark script,
or SparkSQL query, that you submit to Amazon EMR on EKS.
See: https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/emr-containers.html#EMRContainers.Client.start_job_run # noqa: E501
:param name: The name of the job run.
:param execution_role_arn: The IAM role ARN associated with the job run.
:param release_label: The Amazon EMR release version to use for the job run.
:param job_driver: Job configuration details, e.g. the Spark job parameters.
:param configuration_overrides: The configuration overrides for the job run,
specifically either application configuration or monitoring configuration.
:param client_request_token: The client idempotency token of the job run request.
Use this if you want to specify a unique ID to prevent two jobs from getting started.
:param tags: The tags assigned to job runs.
:return: Job ID
"""
params = {
"name": name,
"virtualClusterId": self.virtual_cluster_id,
"executionRoleArn": execution_role_arn,
"releaseLabel": release_label,
"jobDriver": job_driver,
"configurationOverrides": configuration_overrides or {},
"tags": tags or {},
}
if client_request_token:
params["clientToken"] = client_request_token
response = self.conn.start_job_run(**params)
if response['ResponseMetadata']['HTTPStatusCode'] != 200:
raise AirflowException(f'Start Job Run failed: {response}')
else:
self.log.info(
"Start Job Run success - Job Id %s and virtual cluster id %s",
response['id'],
response['virtualClusterId'],
)
return response['id']
def get_job_failure_reason(self, job_id: str) -> str | None:
"""
Fetch the reason for a job failure (e.g. error message). Returns None or reason string.
:param job_id: Id of submitted job run
:return: str
"""
# We absorb any errors if we can't retrieve the job status
reason = None
try:
response = self.conn.describe_job_run(
virtualClusterId=self.virtual_cluster_id,
id=job_id,
)
failure_reason = response['jobRun']['failureReason']
state_details = response["jobRun"]["stateDetails"]
reason = f"{failure_reason} - {state_details}"
except KeyError:
self.log.error('Could not get status of the EMR on EKS job')
except ClientError as ex:
self.log.error('AWS request failed, check logs for more info: %s', ex)
return reason
def check_query_status(self, job_id: str) -> str | None:
"""
Fetch the status of submitted job run. Returns None or one of valid query states.
See: https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/emr-containers.html#EMRContainers.Client.describe_job_run # noqa: E501
:param job_id: Id of submitted job run
:return: str
"""
try:
response = self.conn.describe_job_run(
virtualClusterId=self.virtual_cluster_id,
id=job_id,
)
return response["jobRun"]["state"]
except self.conn.exceptions.ResourceNotFoundException:
# If the job is not found, we raise an exception as something fatal has happened.
raise AirflowException(f'Job ID {job_id} not found on Virtual Cluster {self.virtual_cluster_id}')
except ClientError as ex:
# If we receive a generic ClientError, we swallow the exception so that the
self.log.error('AWS request failed, check logs for more info: %s', ex)
return None
def poll_query_status(
self,
job_id: str,
max_tries: int | None = None,
poll_interval: int = 30,
max_polling_attempts: int | None = None,
) -> str | None:
"""
Poll the status of submitted job run until query state reaches final state.
Returns one of the final states.
:param job_id: Id of submitted job run
:param max_tries: Deprecated - Use max_polling_attempts instead
:param poll_interval: Time (in seconds) to wait between calls to check query status on EMR
:param max_polling_attempts: Number of times to poll for query state before function exits
:return: str
"""
if max_tries:
warnings.warn(
f"Method `{self.__class__.__name__}.max_tries` is deprecated and will be removed "
"in a future release. Please use method `max_polling_attempts` instead.",
DeprecationWarning,
stacklevel=2,
)
if max_polling_attempts and max_polling_attempts != max_tries:
raise Exception("max_polling_attempts must be the same value as max_tries")
else:
max_polling_attempts = max_tries
try_number = 1
final_query_state = None # Query state when query reaches final state or max_polling_attempts reached
while True:
query_state = self.check_query_status(job_id)
if query_state is None:
self.log.info("Try %s: Invalid query state. Retrying again", try_number)
elif query_state in self.TERMINAL_STATES:
self.log.info("Try %s: Query execution completed. Final state is %s", try_number, query_state)
final_query_state = query_state
break
else:
self.log.info("Try %s: Query is still in non-terminal state - %s", try_number, query_state)
if (
max_polling_attempts and try_number >= max_polling_attempts
): # Break loop if max_polling_attempts reached
final_query_state = query_state
break
try_number += 1
sleep(poll_interval)
return final_query_state
def stop_query(self, job_id: str) -> dict:
"""
Cancel the submitted job_run
:param job_id: Id of submitted job_run
:return: dict
"""
return self.conn.cancel_job_run(
virtualClusterId=self.virtual_cluster_id,
id=job_id,
)
相关信息
相关文章
0
赞
热门推荐
-
2、 - 优质文章
-
3、 gate.io
-
8、 golang
-
9、 openharmony
-
10、 Vue中input框自动聚焦