Skip to content

[data] Refractor progress bar clearer metrics #57094

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
alexeykudinkin merged 13 commits into from
Oct 24, 2025
Merged

[data] Refractor progress bar clearer metrics #57094

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
13 commits
Select commit Hold shift + click to select a range
e8a6d08
[data] Refine estimate for total_num_rows in progress bars
iamjustinhsu Oct 1, 2025
a909536
keep finished + ongoing tasks
iamjustinhsu Oct 7, 2025
9690830
Merge branch 'master' of https://github.com/ray-project/ray into jhsu…
iamjustinhsu Oct 8, 2025
648e5b5
move estimation in output generation
iamjustinhsu Oct 9, 2025
e3f5554
remove print
iamjustinhsu Oct 9, 2025
aa2a6b8
more accurate
iamjustinhsu Oct 9, 2025
f63c0c1
case for num_tasks_finished == 0
iamjustinhsu Oct 10, 2025
444cebb
remove code
iamjustinhsu Oct 10, 2025
8a84a06
comments
iamjustinhsu Oct 10, 2025
e8278df
Merge branch 'master' into jhsu/use-outputs-generated-for-pg
iamjustinhsu Oct 23, 2025
f3d854c
fix
iamjustinhsu Oct 23, 2025
f8ba30a
lint
iamjustinhsu Oct 23, 2025
37bd144
check for None or 0
iamjustinhsu Oct 23, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 7 additions & 11 deletions python/ray/data/_internal/execution/interfaces/physical_operator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -827,27 +827,23 @@ def estimate_total_num_of_blocks(

if (
upstream_op_num_outputs > 0
and metrics.num_inputs_received > 0
and metrics.num_tasks_finished > 0
and metrics.average_num_inputs_per_task
and metrics.average_num_outputs_per_task
and metrics.average_rows_outputs_per_task
):
estimated_num_tasks = total_num_tasks
if estimated_num_tasks is None:
estimated_num_tasks = (
upstream_op_num_outputs
/ metrics.num_inputs_received
* num_tasks_submitted
upstream_op_num_outputs / metrics.average_num_inputs_per_task
)

estimated_num_output_bundles = round(
estimated_num_tasks
* metrics.num_outputs_of_finished_tasks
/ metrics.num_tasks_finished
estimated_num_tasks * metrics.average_num_outputs_per_task
)
estimated_output_num_rows = round(
estimated_num_tasks
* metrics.rows_task_outputs_generated
/ metrics.num_tasks_finished
estimated_num_tasks * metrics.average_rows_outputs_per_task
Comment on lines +837 to +844
Copy link
Member

@bveeramani bveeramani Oct 23, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If these metrics are None, this code will raise a TypeError:

  • average_num_inputs_per_task
  • average_num_outputs_per_task
  • average_rows_outputs_per_task

This can't happen right now because we check num_tasks_finished > 0, and the three metrics aren't None when num_tasks_finished > 0. But, that's an implementation detail, and isn't guaranteed by their interface.

This code would be more robust if replaced this check:

        and metrics.num_inputs_received > 0
        and metrics.num_tasks_finished > 0

With explicit checks that the metrics aren't None.

)

return (
estimated_num_tasks,
estimated_num_output_bundles,
Expand Down
3 changes: 0 additions & 3 deletions python/ray/data/_internal/execution/operators/input_data_buffer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,6 @@ def __init__(
data_context: DataContext,
input_data: Optional[List[RefBundle]] = None,
input_data_factory: Optional[Callable[[int], List[RefBundle]]] = None,
num_output_blocks: Optional[int] = None,
):
"""Create an InputDataBuffer.

Expand All @@ -30,8 +29,6 @@ def __init__(
object to use injestion.
input_data: The list of bundles to output from this operator.
input_data_factory: The factory to get input data, if input_data is None.
num_output_blocks: The number of output blocks. If not specified, progress
Copy link
Contributor Author

@iamjustinhsu iamjustinhsu Oct 1, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

not being used anywhere

bars total will be set based on num output bundles instead.
"""
super().__init__("Input", [], data_context)
if input_data is not None:
Expand Down