airflow-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Work logged] (AIRFLOW-2524) Airflow integration with AWS Sagemaker
Date Wed, 01 Aug 2018 18:40:00 GMT

     [ https://issues.apache.org/jira/browse/AIRFLOW-2524?focusedWorklogId=129904&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-129904
]

ASF GitHub Bot logged work on AIRFLOW-2524:
-------------------------------------------

                Author: ASF GitHub Bot
            Created on: 01/Aug/18 18:39
            Start Date: 01/Aug/18 18:39
    Worklog Time Spent: 10m 
      Work Description: srrajeev-aws commented on a change in pull request #3658: [AIRFLOW-2524]
Add Amazon SageMaker Training
URL: https://github.com/apache/incubator-airflow/pull/3658#discussion_r206988684
 
 

 ##########
 File path: airflow/contrib/operators/sagemaker_create_training_job_operator.py
 ##########
 @@ -0,0 +1,98 @@
+# -*- coding: utf-8 -*-
+#
+# 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 airflow.contrib.hooks.sagemaker_hook import SageMakerHook
+from airflow.models import BaseOperator
+from airflow.utils import apply_defaults
+from airflow.exceptions import AirflowException
+
+
+class SageMakerCreateTrainingJobOperator(BaseOperator):
+
+    """
+       Initiate a SageMaker training
+
+       This operator returns The ARN of the model created in Amazon SageMaker
+
+       :param training_job_config:
+       The configuration necessary to start a training job (templated)
+       :type training_job_config: dict
+       :param region_name: The AWS region_name
+       :type region_name: string
+       :param sagemaker_conn_id: The SageMaker connection ID to use.
+       :type aws_conn_id: string
 
 Review comment:
   @Fokko - To further add to Keliang explanation of separating the operator to kick off the
job and sensor to monitor the job is to provide flexibility to users. Based on their use case(s),
they may have the requirements to kick of multiple jobs/tasks in parallel and then monitor
the completion of all Amazon Sagemaker job(s) downstream. Some these jobs may take hours and
we don't want to hold the pipeline to initiate other downstream jobs hampering the users from
meeting their required SLA. Since there are many other known and unknown scenarios, we are
careful not to club both the initialization and monitoring of the job. 
   
   The design is similar to the Amazon EMR - https://github.com/apache/incubator-airflow/blob/master/airflow/contrib/example_dags/example_emr_job_flow_manual_steps.py
        

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Issue Time Tracking
-------------------

            Worklog Id:     (was: 129904)
            Time Spent: 10m
    Remaining Estimate: 0h

> Airflow integration with AWS Sagemaker
> --------------------------------------
>
>                 Key: AIRFLOW-2524
>                 URL: https://issues.apache.org/jira/browse/AIRFLOW-2524
>             Project: Apache Airflow
>          Issue Type: Improvement
>          Components: aws, contrib
>            Reporter: Rajeev Srinivasan
>            Assignee: Yang Yu
>            Priority: Major
>              Labels: AWS
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> Would it be possible to orchestrate an end to end  AWS  Sagemaker job using Airflow.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message