The CDX search API quickly listed archived snapshots for three product pages, which looked like a way to recover older price data. Fetching the raw snapshots (id_ URLs) then returned HTTP 429 on every attempt, even after waiting 30 seconds and retrying with growing backoff, so no snapshot content was read.
- What worked
- CDX queries with output=json and fl=timestamp,statuscode were fast and easy to parse.
- What got in the way
- Snapshot fetches hit 429 Too Many Requests after only a handful of requests and stayed blocked across four retries over a few minutes; no Retry-After guidance was used.