fix: Avoid heavy top-level imports in operators #85
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
For complex DAGs, the import time added by this library could generate `DagBag` import timeouts if the configured value is low enough. An Airflow documented best practice is to [avoid heavy Python code](https://airflow.apache.org/docs/apache-airflow/stable/best-practices.html#top-level-python-code) that runs on DAG and Operator creation, and `dbt` imports are slow, based on profiling. Profiling can be easily run locally, with the following command: ```shell python -X importtime -c "from airflow_dbt_python.operators.dbt import DbtRunOperator" 2>import-times.log ``` And then parsed using a tool like [tuna](https://github.com/nschloe/tuna). Before this change, the operator import takes ~1.37s, which is reduced to ~0.25s with this fix. It's important to note that, from those 0.25s, more than 80% of the time is spent importing `airflow` components, which will be commonly already loaded in DAGs, so this library's import time for operators becomes insignificant.
|
Changes look neat, thanks for your contribution.
Fun story: sometimes dbt does network calls (!) when importing modules. This was a particular headache for me when I was trying to fetch the dbt version fast. Anyways, if CI is green, I'll be happy to merge this and push out a patch release. |
tomasfarias
approved these changes
Nov 22, 2022
|
Thank you for the quick review and release! You have done an excellent job building and supporting this project :) |
adamantike
deleted the
fix/avoid-heavy-top-level-imports-in-operators
branch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
For complex DAGs, the import time added by this library could generate
DagBagimport timeouts if the configured value is low enough. An Airflow documented best practice is toavoid heavy Python code that runs on DAG and Operator creation, and
dbtimports are slow, based on profiling.Profiling can be easily run locally, with the following command:
And then parsed using a tool like tuna.
Before this change, the operator import takes ~1.37s, which is reduced to ~0.25s with this fix.
It's important to note that, from those 0.25s, more than 80% of the time is spent importing
airflowcomponents, which will be commonly already loaded in DAGs, so this library's import time for operators becomes insignificant.