One-time requests

One-time requests are available on every plan. They create one run and retry retryable delivery failures within that run.

Create a request

Set schedule_type to one_time, add a future run_at, and choose an IANA timezone.

curl --fail-with-body -sS -X POST "https://api.cronbolt.com/api/v1/jobs" \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $CRONBOLT_API_KEY" \
  -H "Idempotency-Key: launch-callback-v1" \
  -d '{
    "name": "Send launch callback",
    "schedule_type": "one_time",
    "run_at": "2035-01-01T09:00:00",
    "timezone": "Asia/Kolkata",
    "method": "POST",
    "url": "https://example.com/webhooks/launch",
    "timeout_seconds": 30
  }'

A date without a UTC offset is read in the selected timezone. A date with an offset is converted to UTC before it is stored.

Lifecycle

1. The request remains active while it waits for run_at.
2. Cronbolt creates its single run when it becomes due.
3. The request becomes expired and no longer counts toward the pending limit.
4. The worker delivers the request and applies the saved retry policy.
5. The run records success or failure.

One-time requests cannot be paused, resumed, or run manually. New requests retry retryable failures twice by default. Set max_retries to 0 to disable automatic retries.

Limit

Spark can have 1 pending one-time request. Bolt can have 100, and Surge can have 500. These limits are shared across the account.

Expired one-time requests do not count toward active recurring-job or per-project storage limits.