hankehly commented on code in PR #27076:
URL: https://github.com/apache/airflow/pull/27076#discussion_r996536928
##########
airflow/providers/amazon/aws/operators/rds.py:
##########
@@ -672,6 +682,132 @@ def execute(self, context: Context) -> str:
return json.dumps(delete_db_instance, default=str)
+class RdsStartDbOperator(RdsBaseOperator):
+ """
+ Starts an RDS DB instance / cluster
+
+ .. seealso::
+ For more information on how to use this operator, take a look at the
guide:
+ :ref:`howto/operator:RdsStartDbOperator`
+
+ :param db_identifier: The AWS identifier of the DB to start
+ :param db_type: Type of the DB - either "instance" or "cluster" (default:
"instance")
+ :param aws_conn_id: The Airflow connection used for AWS credentials.
(default: "aws_default")
+ :param wait_for_completion: If True, waits for DB to start. (default:
True)
+ """
+
+ template_fields = ("db_identifier", "db_type")
+
+ def __init__(
+ self,
+ *,
+ db_identifier: str,
+ db_type: str = "instance",
+ aws_conn_id: str = "aws_default",
+ wait_for_completion: bool = True,
+ **kwargs,
+ ):
+ super().__init__(aws_conn_id=aws_conn_id, **kwargs)
+ self.db_identifier = db_identifier
+ self.db_type = db_type
+ self.wait_for_completion = wait_for_completion
+
+ def execute(self, context: Context) -> str:
+ self.db_type = RdsDbType(self.db_type)
+ start_db_response = self._start_db()
+ if self.wait_for_completion:
+ self._wait_until_db_available()
+ return json.dumps(start_db_response, default=str)
+
+ def _start_db(self):
+ self.log.info(f"Starting DB {self.db_type} '{self.db_identifier}'")
+ if self.db_type == RdsDbType.INSTANCE:
+ response =
self.hook.conn.start_db_instance(DBInstanceIdentifier=self.db_identifier)
+ else:
+ response =
self.hook.conn.start_db_cluster(DBClusterIdentifier=self.db_identifier)
+ return response
+
+ def _wait_until_db_available(self):
Review Comment:
@kazanzhy
Thanks for your review. I agree, that would be more consistent. What
concerns me is that when using `_await_status` we need to specify all possible
`wait_statuses`. I observed the following states when creating an RDS instance,
but [depending on the settings, there may be
more.](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/accessing-monitoring.html)
(And they might change in the future?)
- Creating
- Backing-up
- Configuring-enhanced-monitoring
Rather than listing every possible state or risk "missing one," I think
using the builtin "waiter" implementation is more appropriate here. We only
need to check for the "available" state.
@o-nikolas
I noticed you work at Amazon, is there anything you'd like to add to this
discussion?
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
