[jira] [Commented] (AIRFLOW-2777) dag.sub_dag(...) speedups
[ https://issues.apache.org/jira/browse/AIRFLOW-2777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16601286#comment-16601286 ] Apache Spark commented on AIRFLOW-2777: --- User 'abdul-stripe' has created a pull request for this issue: https://github.com/apache/incubator-airflow/pull/3621 > dag.sub_dag(...) speedups > - > > Key: AIRFLOW-2777 > URL: https://issues.apache.org/jira/browse/AIRFLOW-2777 > Project: Apache Airflow > Issue Type: Improvement >Reporter: Abdul Nimeri >Assignee: Abdul Nimeri >Priority: Minor > Fix For: 2.0.0 > > > dag.sub_dag(...) currently works by first deep copying the entire dag, and > then filtering down to the appropriate tasks. that can be slow since deep > copying a big dag takes a while. specifically, copying over all the tasks is > the bottleneck > this can be a lot faster by instead only copying over the filtered down tasks > > -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (AIRFLOW-2777) dag.sub_dag(...) speedups
[ https://issues.apache.org/jira/browse/AIRFLOW-2777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16558761#comment-16558761 ] ASF subversion and git services commented on AIRFLOW-2777: -- Commit 4cd72b91bd024e6be9abb30aee345c0a90a6561b in incubator-airflow's branch refs/heads/master from [~abdul-stripe] [ https://git-wip-us.apache.org/repos/asf?p=incubator-airflow.git;h=4cd72b9 ] [AIRFLOW-2777] speed up dag.sub_dag(...) previous version created the subdag by copying over all the tasks, and then filtering them down. it's a lot faster if we only copy over the tasks we need Closes #3621 from abdul-stripe/faster-subdag > dag.sub_dag(...) speedups > - > > Key: AIRFLOW-2777 > URL: https://issues.apache.org/jira/browse/AIRFLOW-2777 > Project: Apache Airflow > Issue Type: Improvement >Reporter: Abdul Nimeri >Assignee: Abdul Nimeri >Priority: Minor > Fix For: 2.0.0 > > > dag.sub_dag(...) currently works by first deep copying the entire dag, and > then filtering down to the appropriate tasks. that can be slow since deep > copying a big dag takes a while. specifically, copying over all the tasks is > the bottleneck > this can be a lot faster by instead only copying over the filtered down tasks > > -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (AIRFLOW-2777) dag.sub_dag(...) speedups
[ https://issues.apache.org/jira/browse/AIRFLOW-2777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16551422#comment-16551422 ] Abdul Nimeri commented on AIRFLOW-2777: --- https://github.com/apache/incubator-airflow/pull/3621 > dag.sub_dag(...) speedups > - > > Key: AIRFLOW-2777 > URL: https://issues.apache.org/jira/browse/AIRFLOW-2777 > Project: Apache Airflow > Issue Type: Improvement >Reporter: Abdul Nimeri >Assignee: Abdul Nimeri >Priority: Minor > > dag.sub_dag(...) currently works by first deep copying the entire dag, and > then filtering down to the appropriate tasks. that can be slow since deep > copying a big dag takes a while. specifically, copying over all the tasks is > the bottleneck > this can be a lot faster by instead only copying over the filtered down tasks > > -- This message was sent by Atlassian JIRA (v7.6.3#76005)