Skip to content

Add custom retry callback to the Pathways elastic retry manager. #235

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
copybara-service wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add custom retry callback to the Pathways elastic retry manager. #235

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
7 changes: 6 additions & 1 deletion pathwaysutils/elastic/manager.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -197,6 +197,7 @@ def elastic_retry(
timeout: float | None = None,
pre_callback: Callable[..., Any] | None = None,
on_elastic_event_callback: Callable[..., Any] | None = None,
should_retry: Callable[[jax.errors.JaxRuntimeError], bool] | None = None,
) -> Callable[[_F], _F]:
"""Retries a function with elasticity fault tolerance.

Expand Down Expand Up @@ -233,6 +234,7 @@ def elastic_retry(
pre_callback: A callback to call before the function is attempted.
on_elastic_event_callback: A callback to call after an elastic failure
occurs.
should_retry: Custom callback to determine if an error should be retried.

Returns:
A decorator that retries the wrapped function.
Expand Down Expand Up @@ -299,7 +301,10 @@ def attempt_execution(retry_index: int) -> Any:
if on_elastic_event_callback is not None:
on_elastic_event_callback()
except jax.errors.JaxRuntimeError as error:
if not elastic.is_error_due_to_slice_down(error):
should_retry_error = elastic.is_error_due_to_slice_down(error)
if should_retry is not None and should_retry(error):
should_retry_error = True
if not should_retry_error:
raise

if self.new_slice_event.is_set():
Expand Down
Loading