airflow.operators.hive_operator

Module Contents

class airflow.operators.hive_operator.HiveOperator(hql, hive_cli_conn_id='hive_cli_default', schema='default', hiveconfs=None, hiveconf_jinja_translate=False, script_begin_tag=None, run_as_owner=False, mapred_queue=None, mapred_queue_priority=None, mapred_job_name=None, *args, **kwargs)[source]

Bases: airflow.models.BaseOperator

Executes hql code or hive script in a specific Hive database.

Parameters
  • hql (str) – the hql to be executed. Note that you may also use a relative path from the dag file of a (template) hive script. (templated)

  • hive_cli_conn_id (str) – reference to the Hive database. (templated)

  • hiveconfs (dict) – if defined, these key value pairs will be passed to hive as -hiveconf "key"="value"

  • hiveconf_jinja_translate (bool) – when True, hiveconf-type templating ${var} gets translated into jinja-type templating {{ var }} and ${hiveconf:var} gets translated into jinja-type templating {{ var }}. Note that you may want to use this along with the DAG(user_defined_macros=myargs) parameter. View the DAG object documentation for more details.

  • script_begin_tag (str) – If defined, the operator will get rid of the part of the script before the first occurrence of script_begin_tag

  • mapred_queue (str) – queue used by the Hadoop CapacityScheduler. (templated)

  • mapred_queue_priority (str) – priority within CapacityScheduler queue. Possible settings include: VERY_HIGH, HIGH, NORMAL, LOW, VERY_LOW

  • mapred_job_name (str) – This name will appear in the jobtracker. This can make monitoring easier.

template_fields = ['hql', 'schema', 'hive_cli_conn_id', 'mapred_queue', 'hiveconfs', 'mapred_job_name', 'mapred_queue_priority'][source]
template_ext = ['.hql', '.sql'][source]
ui_color = #f0e4ec[source]
get_hook(self)[source]
prepare_template(self)[source]
execute(self, context)[source]
dry_run(self)[source]
on_kill(self)[source]

Was this entry helpful?