This is an automated email from the ASF dual-hosted git repository. ephraimanierobi pushed a commit to branch v2-7-test in repository https://gitbox.apache.org/repos/asf/airflow.git
commit abfe7b89ef9d2f0b0ce46d146b2fd5d595dcfd08 Author: Vijayasarathi Balasubramanian <[email protected]> AuthorDate: Fri Aug 4 15:14:43 2023 -0400 Documentation Update to enhance Readability (#32832) * Documentation Update to enhance Readability Update to plugin.rst to enhance readability * Update docs/apache-airflow/authoring-and-scheduling/plugins.rst --------- Co-authored-by: eladkal <[email protected]> (cherry picked from commit aedefd6516101b182bc9ab3a6740ac5c0a8e2f9d) --- docs/apache-airflow/authoring-and-scheduling/plugins.rst | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/apache-airflow/authoring-and-scheduling/plugins.rst b/docs/apache-airflow/authoring-and-scheduling/plugins.rst index 2a9314f39e..f0c9f43ddc 100644 --- a/docs/apache-airflow/authoring-and-scheduling/plugins.rst +++ b/docs/apache-airflow/authoring-and-scheduling/plugins.rst @@ -82,9 +82,9 @@ start of each Airflow process, set ``[core] lazy_load_plugins = False`` in ``air This means that if you make any changes to plugins and you want the webserver or scheduler to use that new code you will need to restart those processes. However, it will not be reflected in new running tasks after the scheduler boots. -By default, task execution will use forking to avoid the slow down of having to create a whole new python -interpreter and re-parse all of the Airflow code and start up routines -- this is a big benefit for shorter -running tasks. This does mean that if you use plugins in your tasks, and want them to update you will either +By default, task execution uses forking. This avoids the slowdown associated with creating a new Python interpreter +and re-parsing all of Airflow's code and startup routines. This approach offers significant benefits, especially for shorter tasks. +This does mean that if you use plugins in your tasks, and want them to update you will either need to restart the worker (if using CeleryExecutor) or scheduler (Local or Sequential executors). The other option is you can accept the speed hit at start up set the ``core.execute_tasks_new_python_interpreter`` config setting to True, resulting in launching a whole new python interpreter for tasks.
