Hi ! I’m developing a non-commercial radiation mapping project using the Safecast API to fetch measurements and display them on a heatmap via Firestore. My script queries `/measurements.json?since=...&per_page=1000` every 2 hours through GitHub Actions but consistently encounters 503 Service Unavailable errors (“Back-end server is at capacity”), even with retries (5-40s delays). I’ve reduced `per_page` to 100 and added pagination, but the issue persists.
Additionally, I’m facing challenges with `device_info` mapping. The `/measurements.json` endpoint returns high-numbered `device_id`s (e.g., 100412, 65006), but `/devices` and `/devices/{device_id}.json` return unrelated IDs (e.g., 4-28) or no data, leaving `device_info` empty. I understand from your web map data flow documentation that bGeigies may have NULL `device_id`s, which could explain mismatches.
Could you assist with:
1. Are 503 errors due to free-tier rate limits or server capacity? Would smaller `per_page` (e.g., 50) or specific request timing help?
2. Why do `device_id`s from `/measurements.json` not match `/devices`? Is there a correct endpoint or table for these IDs?
3. Your documentation mentions a filtered dataset and SQLite database for download. Can you provide the URLs for these bulk exports (CSV/JSON or SQLite)?
4. Is a higher-limit API key available for non-commercial projects to improve reliability?
Any guidance would be greatly appreciated.
Thank you,
Mariia Tsymberh