Skip to content

[FLINK-39532][python] Fix race condition in Python AsyncScalarFunctionOperation #28036

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

Open
auroflow wants to merge 1 commit into
base: master
Choose a base branch
from
Open

[FLINK-39532][python] Fix race condition in Python AsyncScalarFunctionOperation #28036

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 11 additions & 0 deletions flink-python/pyflink/fn_execution/table/async_function/operations.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,6 +60,11 @@ def __init__(self, serialized_fn):
operation_utils.extract_user_defined_function(
serialized_fn.udfs[0], one_arg_optimization=False)

# Mirror PythonScalarFunctionOperator.createInputCoderInfoDescriptor:
# Java picks FlattenRowCoder unless some UDF takes a row as input.
self._input_is_flatten_row = not any(
udf.takes_row_as_input for udf in serialized_fn.udfs)

# Create the eval function
self._eval_func = eval('lambda value: %s' % scalar_function, variable_dict)

Expand Down Expand Up @@ -140,6 +145,12 @@ def process_element(self, value):
"""
self._raise_exception_if_exists()

# The Cython FlattenRowCoderImpl returns a reused list whose slots are
# overwritten by the next decode, so we must snapshot the row before
# the async closure can capture it.
if self._input_is_flatten_row:
value = list(value)

entry = self._queue.put(None, 0, 0, value)

async def execute_async(rh):
Expand Down