Black Friday spike: your webhook returns 503, what Sume retries
If your endpoint sheds load on Black Friday, Sume retries a terminal webhook up to 10 times with doubling backoff. Know the schedule and what to do after.

Sume does not drop a terminal webhook on the first failed attempt. It tries up to 10 times, each with a 10 second limit, and the delay starts at 30 seconds, doubles per attempt with jitter and is capped at one hour. A Retry-After header from your endpoint is honored.
What counts as a failed attempt
- A non-success status such as 503 or 429.
- No response within 10 seconds.
- A 3xx redirect. Sume does not follow it.
- A URL that no longer passes public HTTPS validation.
Shed load on purpose
If you must refuse during a spike, return 503 with Retry-After; Sume will come back. Better: acknowledge fast and queue the work. Because a delivery can repeat, de-duplicate on request_id, which equals the run id.
app.post('/hooks/sume', express.raw({ type: '*/*' }), async (req, res) => {
// verify x-sume-webhook-signature against req.body first
const evt = JSON.parse(req.body.toString());
if (await seen(evt.request_id)) return res.sendStatus(200);
await enqueue(evt); // do real work later
res.sendStatus(200);
});After the last attempt
Delivery status moves through pending and retrying to delivered, or to exhausted. For an exhausted run, read it from GET /v1/format-runs/{run_id} and replay it with POST /v1/format-runs/{run_id}/webhook/redeliver once your endpoint is healthy.
Sources
Related posts
More in Developers
- Save a 30-second Seedance 2.5 clip to S3 with boto3, no temp file
Stream the finished MP4 from Sume's /content endpoint straight into S3 with boto3 upload_fileobj. Includes the 30 s price table and a short script.
- One source image, three Ideogram 4.5 edits in parallel (Python)
Run three different Ideogram 4.5 edits from the same source image at once on Sume, each with its own Idempotency-Key, instead of chaining them.
- Budget a week of Shorts before rendering: Timeline plan sum
Sume's unbilled Timeline plan returns billable minutes and an estimated cost. A Python script sums seven plans so you know the week's render cost first.
- Bulgarian, Croatian, Slovak text to speech API: one request each
Cartesia's Sonic 3.6 list includes bg, hr and sk. Run three scripts through Sume TTS 1.0 with a language map. Three 2,000-character jobs quote 30 cents.
Written by Sume