Skip to content

Fix: missing stats #54081

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

Closed
kempspo wants to merge 2 commits into from
Closed

Fix: missing stats #54081

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3717ebc
add stats incr for ti.start and ti.finish
kempspo Jul 28, 2025
b47d890
add stats incr for failures
kempspo Aug 4, 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
34 changes: 33 additions & 1 deletion task-sdk/src/airflow/sdk/execution_time/task_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,6 +105,7 @@
)
from airflow.sdk.execution_time.xcom import XCom
from airflow.sdk.timezone import coerce_datetime
from airflow.stats import Stats

if TYPE_CHECKING:
import jinja2
Expand Down Expand Up @@ -552,6 +553,11 @@ def get_dagrun_state(dag_id: str, run_id: str) -> str:

return response.state

@property
def stats_tags(self) -> dict[str, str]:
"""Returns task instance tags."""
return {"dag_id": self.dag_id, "task_id": self.task_id}


def _xcom_push(ti: RuntimeTaskInstance, key: str, value: Any, mapped_length: int | None = None) -> None:
"""Push a XCom through XCom.set, which pushes to XCom Backend if configured."""
Expand Down Expand Up @@ -893,7 +899,21 @@ def _on_term(signum, frame):
msg: ToSupervisor | None = None
state: TaskInstanceState
error: BaseException | None = None

Stats.incr(f"ti.start.{ti.dag_id}.{ti.task_id}", tags=ti.stats_tags)
# Same metric with tagging
Stats.incr("ti.start", tags=ti.stats_tags)
for state in TaskInstanceState:
Stats.incr(
f"ti.finish.{ti.dag_id}.{ti.task_id}.{state.value}",
count=0,
tags=ti.stats_tags,
)
# Same metric with tagging
Stats.incr(
"ti.finish",
count=0,
tags={**ti.stats_tags, "state": str(state.value)},
)
try:
# First, clear the xcom data sent from server
if ti._ti_context_from_server and (keys_to_delete := ti._ti_context_from_server.xcom_keys_to_clear):
Expand Down Expand Up @@ -991,6 +1011,12 @@ def _on_term(signum, frame):
msg, state = _handle_current_task_failed(ti)
error = e
finally:
Stats.incr(
f"ti.finish.{ti.dag_id}.{ti.task_id}.{ti.state}",
tags=ti.stats_tags,
)
# Same metric with tagging
Stats.incr("ti.finish", tags={**ti.stats_tags, "state": str(ti.state)})
if msg:
SUPERVISOR_COMMS.send(msg=msg)

Expand Down Expand Up @@ -1020,6 +1046,12 @@ def _handle_current_task_failed(
end_date = datetime.now(tz=timezone.utc)
if ti._ti_context_from_server and ti._ti_context_from_server.should_retry:
return RetryTask(end_date=end_date), TaskInstanceState.UP_FOR_RETRY

Stats.incr(f"operator_failures_{ti.task}", tags=ti.stats_tags)
# Same metric with tagging
Stats.incr("operator_failures", tags={**ti.stats_tags, "operator": ti.task})
Stats.incr("ti_failures", tags=ti.stats_tags)

return TaskState(
state=TaskInstanceState.FAILED, end_date=end_date, rendered_map_index=ti.rendered_map_index
), TaskInstanceState.FAILED
Expand Down
Loading