Summary
df.status() accepts an instance_id, not a durable function label. This is easy to misuse when callers start workflows with labels and later try to verify the run by passing the label directly to df.status().
Expected behavior
There should be a clear and ergonomic way to check status for a labeled run. At minimum, docs and examples should show callers how to resolve a label to an instance_id before calling df.status().
Possible options:
- Document the pattern of using
df.list_instances() filtered by label, then passing the selected instance_id to df.status().
- Add a helper such as
df.status_by_label(label) or df.latest_instance_for_label(label) if label-based lookup is intended to be a first-class workflow.
Actual behavior
Passing a label directly to df.status() does not return the intended instance status because the argument is interpreted as an instance_id.
Impact
Tests, examples, and clients can appear to verify a labeled workflow while actually querying the wrong identifier.
Summary
df.status()accepts aninstance_id, not a durable function label. This is easy to misuse when callers start workflows with labels and later try to verify the run by passing the label directly todf.status().Expected behavior
There should be a clear and ergonomic way to check status for a labeled run. At minimum, docs and examples should show callers how to resolve a label to an
instance_idbefore callingdf.status().Possible options:
df.list_instances()filtered by label, then passing the selectedinstance_idtodf.status().df.status_by_label(label)ordf.latest_instance_for_label(label)if label-based lookup is intended to be a first-class workflow.Actual behavior
Passing a label directly to
df.status()does not return the intended instance status because the argument is interpreted as aninstance_id.Impact
Tests, examples, and clients can appear to verify a labeled workflow while actually querying the wrong identifier.