Github user JoshRosen commented on a diff in the pull request:
https://github.com/apache/spark/pull/8835#discussion_r40133051
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/pythonUDFs.scala ---
@@ -329,7 +329,13 @@ case class EvaluatePython(
/**
* :: DeveloperApi ::
* Uses PythonRDD to evaluate a [[PythonUDF]], one partition of tuples at
a time.
- * The input data is zipped with the result of the udf evaluation.
+ *
+ * Python evaluation works by sending the necessary (projected) input data
via a socket to an
+ * external Python process, and combine the result from the Python process
with the original row.
+ *
+ * For each row we send to Python, we also put it in a queue. For each
output row from Python,
+ * we drain the queue to find the original input row. Note that if the
Python process is way too
+ * slow, this could lead to the queue growing unbounded and eventually run
out of memory.
--- End diff --
Could we mitigate this by using a
[LinkedBlockingDeque](https://docs.oracle.com/javase/7/docs/api/java/util/concurrent/LinkedBlockingDeque.html)
to have the producer-side block on inserts once the queue grows to a certain
size?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
