andygrove commented on code in PR #2038:
URL:
https://github.com/apache/datafusion-ballista/pull/2038#discussion_r3582563361
##########
ballista/core/src/execution_plans/sort_shuffle/writer.rs:
##########
@@ -195,182 +289,286 @@ impl SortShuffleWriterExec {
.partition_count()
}
- /// Execute the sort-based shuffle write for a single input partition.
+ /// Executes the sort-based shuffle write for this task.
+ ///
+ /// One pipeline per input partition of the
(scheduler-side-shrink-restricted)
Review Comment:
I want to check that I am understanding this correctly. The current model is
one input partition per task, producing one output data file + index, and the
new model ends up producing the same number of files, just moving the
concurrency within a task rather than across tasks?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
