Skip to main content

run_specific_job

Function run_specific_job 

Source
pub fn run_specific_job(
    pool: &Pool,
    id: &DecayJobId,
    now: DateTime<Utc>,
    summary_backend: &dyn SummaryBackend,
) -> DecayResult<()>
Expand description

Run a specific decay job by id.

Idempotency: a terminal job (Completed, Failed, Cancelled) is a no-op. A Pending job is dispatched. An InProgress job is treated as a recovery situation — the runner refuses to re-claim it (another runner may still be live) and surfaces a validation error rather than silently re-running.

summary_backend is consulted only for LLM-summary jobs. See run_next_pending_job for the backend posture.