Skip to content

fix(scheduler): retry failed nodes when resuming a job #146

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

Draft
SavinRazvan wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

fix(scheduler): retry failed nodes when resuming a job #146

Show file tree
Hide file tree
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
4 changes: 2 additions & 2 deletions src/core/scheduler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,8 +109,8 @@ async def execute(
result.outcomes[checkpoint.node_id] = outcome
if checkpoint.status == CheckpointStatus.COMPLETED:
completed.add(checkpoint.node_id)
elif checkpoint.status == CheckpointStatus.FAILED:
failed.add(checkpoint.node_id)
# FAILED/RUNNING/CANCELLED checkpoints are not terminal on resume;
# handlers are re-executed so operators can recover failed jobs.

base_payload = dict(initial_payload or {})
while True:
Expand Down
14 changes: 9 additions & 5 deletions tests/modules/core/test_scheduler_checkpoint_resume.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,12 +55,15 @@ async def node_b(payload: dict) -> dict:
assert calls == {"a": 1, "b": 1}


def test_scheduler_resume_marks_failed_checkpoint_nodes() -> None:
async def node_ok(_: dict) -> dict:
def test_scheduler_resume_retries_failed_checkpoint_nodes() -> None:
calls = {"fragile": 0}

async def node_recoverable(_: dict) -> dict:
calls["fragile"] += 1
return {"ok": True}

async def _scenario() -> None:
graph = TaskGraph([TaskNode(node_id="fragile", handler=node_ok)])
graph = TaskGraph([TaskNode(node_id="fragile", handler=node_recoverable)])
checkpoints = InMemoryCheckpointStore()
scheduler = TaskScheduler(worker_pool=WorkerPool(max_concurrency=1), checkpoint_store=checkpoints)
await checkpoints.save_checkpoint(
Expand All @@ -75,8 +78,9 @@ async def _scenario() -> None:
)
result = await scheduler.execute(job_id="job_failed_ckpt", graph=graph, resume=True)
assert "fragile" in result.outcomes
assert result.outcomes["fragile"].status == TaskStatus.FAILED
assert result.failed is True
assert result.outcomes["fragile"].status == TaskStatus.COMPLETED
assert calls["fragile"] == 1
assert result.failed is False

asyncio.run(_scenario())

Expand Down