Source code for airflow.operators.sensors

# -*- coding: utf-8 -*-
# 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
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import print_function
from future import standard_library
from builtins import str
from past.builtins import basestring

from datetime import datetime
import logging
from urllib.parse import urlparse
from time import sleep
import re
import sys

import airflow
from airflow import hooks, settings
from airflow.exceptions import AirflowException, AirflowSensorTimeout, AirflowSkipException
from airflow.models import BaseOperator, TaskInstance
from airflow.hooks.base_hook import BaseHook
from airflow.hooks.hdfs_hook import HDFSHook
from airflow.utils.state import State
from airflow.utils.decorators import apply_defaults

[docs]class BaseSensorOperator(BaseOperator): ''' Sensor operators are derived from this class an inherit these attributes. Sensor operators keep executing at a time interval and succeed when a criteria is met and fail if and when they time out. :param soft_fail: Set to true to mark the task as SKIPPED on failure :type soft_fail: bool :param poke_interval: Time in seconds that the job should wait in between each tries :type poke_interval: int :param timeout: Time, in seconds before the task times out and fails. :type timeout: int ''' ui_color = '#e6f1f2' @apply_defaults def __init__( self, poke_interval=60, timeout=60*60*24*7, soft_fail=False, *args, **kwargs): super(BaseSensorOperator, self).__init__(*args, **kwargs) self.poke_interval = poke_interval self.soft_fail = soft_fail self.timeout = timeout def poke(self, context): ''' Function that the sensors defined while deriving this class should override. ''' raise AirflowException('Override me.') def execute(self, context): started_at = while not self.poke(context): if ( - started_at).total_seconds() > self.timeout: if self.soft_fail: raise AirflowSkipException('Snap. Time is OUT.') else: raise AirflowSensorTimeout('Snap. Time is OUT.') sleep(self.poke_interval)"Success criteria met. Exiting.")
class SqlSensor(BaseSensorOperator): """ Runs a sql statement until a criteria is met. It will keep trying until sql returns no row, or if the first cell in (0, '0', ''). :param conn_id: The connection to run the sensor against :type conn_id: string :param sql: The sql to run. To pass, it needs to return at least one cell that contains a non-zero / empty string value. """ template_fields = ('sql',) template_ext = ('.hql', '.sql',) ui_color = '#7c7287' @apply_defaults def __init__(self, conn_id, sql, *args, **kwargs): self.sql = sql self.conn_id = conn_id super(SqlSensor, self).__init__(*args, **kwargs) def poke(self, context): hook = BaseHook.get_connection(self.conn_id).get_hook()'Poking: ' + self.sql) records = hook.get_records(self.sql) if not records: return False else: if str(records[0][0]) in ('0', '',): return False else: return True print(records[0][0]) class MetastorePartitionSensor(SqlSensor): """ An alternative to the HivePartitionSensor that talk directly to the MySQL db. This was created as a result of observing sub optimal queries generated by the Metastore thrift service when hitting subpartitioned tables. The Thrift service's queries were written in a way that wouldn't leverage the indexes. :param schema: the schema :type schema: str :param table: the table :type table: str :param partition_name: the partition name, as defined in the PARTITIONS table of the Metastore. Order of the fields does matter. Examples: ``ds=2016-01-01`` or ``ds=2016-01-01/sub=foo`` for a sub partitioned table :type partition_name: str :param mysql_conn_id: a reference to the MySQL conn_id for the metastore :type mysql_conn_id: str """ template_fields = ('partition_name', 'table', 'schema') ui_color = '#8da7be' @apply_defaults def __init__( self, table, partition_name, schema="default", mysql_conn_id="metastore_mysql", *args, **kwargs): self.partition_name = partition_name self.table = table self.schema = schema self.first_poke = True self.conn_id = mysql_conn_id # TODO(aoen): We shouldn't be using SqlSensor here but MetastorePartitionSensor. # The problem is the way apply_defaults works isn't compatible with inheritance. # The inheritance model needs to be reworked in order to support overriding args/ # kwargs with arguments here, then 'conn_id' and 'sql' can be passed into the # constructor below and apply_defaults will no longer throw an exception. super(SqlSensor, self).__init__(*args, **kwargs) def poke(self, context): if self.first_poke: self.first_poke = False if '.' in self.table: self.schema, self.table = self.table.split('.') self.sql = """ SELECT 'X' FROM PARTITIONS A0 LEFT OUTER JOIN TBLS B0 ON A0.TBL_ID = B0.TBL_ID LEFT OUTER JOIN DBS C0 ON B0.DB_ID = C0.DB_ID WHERE B0.TBL_NAME = '{self.table}' AND C0.NAME = '{self.schema}' AND A0.PART_NAME = '{self.partition_name}'; """.format(self=self) return super(MetastorePartitionSensor, self).poke(context) class ExternalTaskSensor(BaseSensorOperator): """ Waits for a task to complete in a different DAG :param external_dag_id: The dag_id that contains the task you want to wait for :type external_dag_id: string :param external_task_id: The task_id that contains the task you want to wait for :type external_task_id: string :param allowed_states: list of allowed states, default is ``['success']`` :type allowed_states: list :param execution_delta: time difference with the previous execution to look at, the default is the same execution_date as the current task. For yesterday, use [positive!] datetime.timedelta(days=1). Either execution_delta or execution_date_fn can be passed to ExternalTaskSensor, but not both. :type execution_delta: datetime.timedelta :param execution_date_fn: function that receives the current execution date and returns the desired execution date to query. Either execution_delta or execution_date_fn can be passed to ExternalTaskSensor, but not both. :type execution_date_fn: callable """ ui_color = '#19647e' @apply_defaults def __init__( self, external_dag_id, external_task_id, allowed_states=None, execution_delta=None, execution_date_fn=None, *args, **kwargs): super(ExternalTaskSensor, self).__init__(*args, **kwargs) self.allowed_states = allowed_states or [State.SUCCESS] if execution_delta is not None and execution_date_fn is not None: raise ValueError( 'Only one of `execution_date` or `execution_date_fn` may' 'be provided to ExternalTaskSensor; not both.') self.execution_delta = execution_delta self.execution_date_fn = execution_date_fn self.external_dag_id = external_dag_id self.external_task_id = external_task_id def poke(self, context): if self.execution_delta: dttm = context['execution_date'] - self.execution_delta elif self.execution_date_fn: dttm = self.execution_date_fn(context['execution_date']) else: dttm = context['execution_date'] 'Poking for ' '{self.external_dag_id}.' '{self.external_task_id} on ' '{dttm} ... '.format(**locals())) TI = TaskInstance session = settings.Session() count = session.query(TI).filter( TI.dag_id == self.external_dag_id, TI.task_id == self.external_task_id, TI.state.in_(self.allowed_states), TI.execution_date == dttm, ).count() session.commit() session.close() return count class NamedHivePartitionSensor(BaseSensorOperator): """ Waits for a set of partitions to show up in Hive. :param partition_names: List of fully qualified names of the partitions to wait for. A fully qualified name is of the form ``schema.table/pk1=pv1/pk2=pv2``, for example, default.users/ds=2016-01-01. This is passed as is to the metastore Thrift client ``get_partitions_by_name`` method. Note that you cannot use logical or comparison operators as in HivePartitionSensor. :type partition_names: list of strings :param metastore_conn_id: reference to the metastore thrift service connection id :type metastore_conn_id: str """ template_fields = ('partition_names', ) ui_color = '#8d99ae' @apply_defaults def __init__( self, partition_names, metastore_conn_id='metastore_default', poke_interval=60 * 3, *args, **kwargs): super(NamedHivePartitionSensor, self).__init__( poke_interval=poke_interval, *args, **kwargs) if isinstance(partition_names, basestring): raise TypeError('partition_names must be an array of strings') self.metastore_conn_id = metastore_conn_id self.partition_names = partition_names self.next_poke_idx = 0 @classmethod def parse_partition_name(self, partition): try: schema, table_partition = partition.split('.', 1) table, partition = table_partition.split('/', 1) return schema, table, partition except ValueError as e: raise ValueError('Could not parse ' + partition) def poke(self, context): if not hasattr(self, 'hook'): self.hook = hooks.HiveMetastoreHook( metastore_conn_id=self.metastore_conn_id) def poke_partition(partition): schema, table, partition = self.parse_partition_name(partition) 'Poking for {schema}.{table}/{partition}'.format(**locals()) ) return self.hook.check_for_named_partition( schema, table, partition) while self.next_poke_idx < len(self.partition_names): if poke_partition(self.partition_names[self.next_poke_idx]): self.next_poke_idx += 1 else: return False return True class HivePartitionSensor(BaseSensorOperator): """ Waits for a partition to show up in Hive. Note: Because ``partition`` supports general logical operators, it can be inefficient. Consider using NamedHivePartitionSensor instead if you don't need the full flexibility of HivePartitionSensor. :param table: The name of the table to wait for, supports the dot notation (my_database.my_table) :type table: string :param partition: The partition clause to wait for. This is passed as is to the metastore Thrift client ``get_partitions_by_filter`` method, and apparently supports SQL like notation as in ``ds='2015-01-01' AND type='value'`` and comparison operators as in ``"ds>=2015-01-01"`` :type partition: string :param metastore_conn_id: reference to the metastore thrift service connection id :type metastore_conn_id: str """ template_fields = ('schema', 'table', 'partition',) ui_color = '#2b2d42' @apply_defaults def __init__( self, table, partition="ds='{{ ds }}'", metastore_conn_id='metastore_default', schema='default', poke_interval=60*3, *args, **kwargs): super(HivePartitionSensor, self).__init__( poke_interval=poke_interval, *args, **kwargs) if not partition: partition = "ds='{{ ds }}'" self.metastore_conn_id = metastore_conn_id self.table = table self.partition = partition self.schema = schema def poke(self, context): if '.' in self.table: self.schema, self.table = self.table.split('.') 'Poking for table {self.schema}.{self.table}, ' 'partition {self.partition}'.format(**locals())) if not hasattr(self, 'hook'): self.hook = hooks.HiveMetastoreHook( metastore_conn_id=self.metastore_conn_id) return self.hook.check_for_partition( self.schema, self.table, self.partition) class HdfsSensor(BaseSensorOperator): """ Waits for a file or folder to land in HDFS """ template_fields = ('filepath',) ui_color = settings.WEB_COLORS['LIGHTBLUE'] @apply_defaults def __init__( self, filepath, hdfs_conn_id='hdfs_default', ignored_ext=['_COPYING_'], ignore_copying=True, file_size=None, hook=HDFSHook, *args, **kwargs): super(HdfsSensor, self).__init__(*args, **kwargs) self.filepath = filepath self.hdfs_conn_id = hdfs_conn_id self.file_size = file_size self.ignored_ext = ignored_ext self.ignore_copying = ignore_copying self.hook = hook @staticmethod def filter_for_filesize(result, size=None): """ Will test the filepath result and test if its size is at least self.filesize :param result: a list of dicts returned by Snakebite ls :param size: the file size in MB a file should be at least to trigger True :return: (bool) depending on the matching criteria """ if size: logging.debug('Filtering for file size >= %s in files: %s', size, map(lambda x: x['path'], result)) size *= settings.MEGABYTE result = [x for x in result if x['length'] >= size] logging.debug('HdfsSensor.poke: after size filter result is %s', result) return result @staticmethod def filter_for_ignored_ext(result, ignored_ext, ignore_copying): """ Will filter if instructed to do so the result to remove matching criteria :param result: (list) of dicts returned by Snakebite ls :param ignored_ext: (list) of ignored extentions :param ignore_copying: (bool) shall we ignore ? :return: """ if ignore_copying: regex_builder = "^.*\.(%s$)$" % '$|'.join(ignored_ext) ignored_extentions_regex = re.compile(regex_builder) logging.debug('Filtering result for ignored extentions: %s in files %s', ignored_extentions_regex.pattern, map(lambda x: x['path'], result)) result = [x for x in result if not ignored_extentions_regex.match(x['path'])] logging.debug('HdfsSensor.poke: after ext filter result is %s', result) return result def poke(self, context): sb = self.hook(self.hdfs_conn_id).get_conn() logging.getLogger("snakebite").setLevel(logging.WARNING)'Poking for file {self.filepath} '.format(**locals())) try: # IMOO it's not right here, as there no raise of any kind. # if the filepath is let's say '/data/mydirectory', it's correct but if it is '/data/mydirectory/*', # it's not correct as the directory exists and sb does not raise any error # here is a quick fix result = [f for f in[self.filepath], include_toplevel=False)] logging.debug('HdfsSensor.poke: result is %s', result) result = self.filter_for_ignored_ext(result, self.ignored_ext, self.ignore_copying) result = self.filter_for_filesize(result, self.file_size) return bool(result) except: e = sys.exc_info() logging.debug("Caught an exception !: %s", str(e)) return False class WebHdfsSensor(BaseSensorOperator): """ Waits for a file or folder to land in HDFS """ template_fields = ('filepath',) @apply_defaults def __init__( self, filepath, webhdfs_conn_id='webhdfs_default', *args, **kwargs): super(WebHdfsSensor, self).__init__(*args, **kwargs) self.filepath = filepath self.webhdfs_conn_id = webhdfs_conn_id def poke(self, context): c = airflow.hooks.webhdfs_hook.WebHDFSHook(self.webhdfs_conn_id) 'Poking for file {self.filepath} '.format(**locals())) return c.check_for_path(hdfs_path=self.filepath) class S3KeySensor(BaseSensorOperator): """ Waits for a key (a file-like instance on S3) to be present in a S3 bucket. S3 being a key/value it does not support folders. The path is just a key a resource. :param bucket_key: The key being waited on. Supports full s3:// style url or relative path from root level. :type bucket_key: str :param bucket_name: Name of the S3 bucket :type bucket_name: str :param wildcard_match: whether the bucket_key should be interpreted as a Unix wildcard pattern :type wildcard_match: bool :param s3_conn_id: a reference to the s3 connection :type s3_conn_id: str """ template_fields = ('bucket_key', 'bucket_name') @apply_defaults def __init__( self, bucket_key, bucket_name=None, wildcard_match=False, s3_conn_id='s3_default', *args, **kwargs): super(S3KeySensor, self).__init__(*args, **kwargs) # Parse if bucket_name is None: parsed_url = urlparse(bucket_key) if parsed_url.netloc == '': raise AirflowException('Please provide a bucket_name') else: bucket_name = parsed_url.netloc if parsed_url.path[0] == '/': bucket_key = parsed_url.path[1:] else: bucket_key = parsed_url.path self.bucket_name = bucket_name self.bucket_key = bucket_key self.wildcard_match = wildcard_match self.s3_conn_id = s3_conn_id def poke(self, context): import airflow.hooks.S3_hook hook = airflow.hooks.S3_hook.S3Hook(s3_conn_id=self.s3_conn_id) full_url = "s3://" + self.bucket_name + "/" + self.bucket_key'Poking for key : {full_url}'.format(**locals())) if self.wildcard_match: return hook.check_for_wildcard_key(self.bucket_key, self.bucket_name) else: return hook.check_for_key(self.bucket_key, self.bucket_name) class S3PrefixSensor(BaseSensorOperator): """ Waits for a prefix to exist. A prefix is the first part of a key, thus enabling checking of constructs similar to glob airfl* or SQL LIKE 'airfl%'. There is the possibility to precise a delimiter to indicate the hierarchy or keys, meaning that the match will stop at that delimiter. Current code accepts sane delimiters, i.e. characters that are NOT special characters in the Python regex engine. :param bucket_name: Name of the S3 bucket :type bucket_name: str :param prefix: The prefix being waited on. Relative path from bucket root level. :type prefix: str :param delimiter: The delimiter intended to show hierarchy. Defaults to '/'. :type delimiter: str """ template_fields = ('prefix', 'bucket_name') @apply_defaults def __init__( self, bucket_name, prefix, delimiter='/', s3_conn_id='s3_default', *args, **kwargs): super(S3PrefixSensor, self).__init__(*args, **kwargs) # Parse self.bucket_name = bucket_name self.prefix = prefix self.delimiter = delimiter self.full_url = "s3://" + bucket_name + '/' + prefix self.s3_conn_id = s3_conn_id def poke(self, context):'Poking for prefix : {self.prefix}\n' 'in bucket s3://{self.bucket_name}'.format(**locals())) import airflow.hooks.S3_hook hook = airflow.hooks.S3_hook.S3Hook(s3_conn_id=self.s3_conn_id) return hook.check_for_prefix( prefix=self.prefix, delimiter=self.delimiter, bucket_name=self.bucket_name) class TimeSensor(BaseSensorOperator): """ Waits until the specified time of the day. :param target_time: time after which the job succeeds :type target_time: datetime.time """ template_fields = tuple() @apply_defaults def __init__(self, target_time, *args, **kwargs): super(TimeSensor, self).__init__(*args, **kwargs) self.target_time = target_time def poke(self, context): 'Checking if the time ({0}) has come'.format(self.target_time)) return > self.target_time class TimeDeltaSensor(BaseSensorOperator): """ Waits for a timedelta after the task's execution_date + schedule_interval. In Airflow, the daily task stamped with ``execution_date`` 2016-01-01 can only start running on 2016-01-02. The timedelta here represents the time after the execution period has closed. :param delta: time length to wait after execution_date before succeeding :type delta: datetime.timedelta """ template_fields = tuple() @apply_defaults def __init__(self, delta, *args, **kwargs): super(TimeDeltaSensor, self).__init__(*args, **kwargs) = delta def poke(self, context): dag = context['dag'] target_dttm = dag.following_schedule(context['execution_date']) target_dttm +='Checking if the time ({0}) has come'.format(target_dttm)) return > target_dttm class HttpSensor(BaseSensorOperator): """ Executes a HTTP get statement and returns False on failure: 404 not found or response_check function returned False :param http_conn_id: The connection to run the sensor against :type http_conn_id: string :param endpoint: The relative part of the full url :type endpoint: string :param params: The parameters to be added to the GET url :type params: a dictionary of string key/value pairs :param headers: The HTTP headers to be added to the GET request :type headers: a dictionary of string key/value pairs :param response_check: A check against the 'requests' response object. Returns True for 'pass' and False otherwise. :type response_check: A lambda or defined function. :param extra_options: Extra options for the 'requests' library, see the 'requests' documentation (options to modify timeout, ssl, etc.) :type extra_options: A dictionary of options, where key is string and value depends on the option that's being modified. """ template_fields = ('endpoint',) @apply_defaults def __init__(self, endpoint, http_conn_id='http_default', params=None, headers=None, response_check=None, extra_options=None, *args, **kwargs): super(HttpSensor, self).__init__(*args, **kwargs) self.endpoint = endpoint self.http_conn_id = http_conn_id self.params = params or {} self.headers = headers or {} self.extra_options = extra_options or {} self.response_check = response_check self.hook = hooks.http_hook.HttpHook(method='GET', http_conn_id=http_conn_id) def poke(self, context):'Poking: ' + self.endpoint) try: response =, data=self.params, headers=self.headers, extra_options=self.extra_options) if self.response_check: # run content check on response return self.response_check(response) except AirflowException as ae: if str(ae).startswith("404"): return False raise ae return True