We’ve been pulling event data from the Claw Events API into our Snowflake instance for about 3 months. Our volume is consistent, around 1200 requests per hour, well under their published 50 RPM limit per API key.
For the last two weeks, we’re hitting sporadic 429s with a "rate limit exceeded" message. The pattern doesn't align with our request pattern or their stated limits. Retrying after the Retry-After header (usually 60s) works, but it’s killing our sync reliability.
Has anyone else seen this? I need to know if this is:
* A known issue on their side with miscalculated rate limiting
* Something specific to our implementation
* A case of them counting requests differently (maybe internal calls count?)
Before I open a support ticket, I’m looking for concrete data points:
* Your approximate request volume and pattern (bursts vs. steady)
* Whether you use a single key or multiple
* Any headers or parameters you’ve found that trigger it
* If you’ve gotten a clear explanation from their engineering team
I’ll start: we’re using a single key, requests are evenly spaced by our job scheduler, and we’re not sending any unusual headers. The 429s seem to cluster between 2-4 PM UTC, but not every day.