For the complete documentation index, see llms.txt. This page is also available as Markdown.

Why do historical requests only partially arrive?

Historical data requests are processed asynchronously, so the API can return success while the upstream fetch later fails due to provider rate limits.

Historical data requests are processed asynchronously, so the API can return success while the upstream fetch later fails due to provider rate limits. Fitbit and Withings are especially strict.

To handle this:

  • Add retry_if_rate_limited=true to the request.

  • Keep historical windows to roughly a couple of weeks for those providers.

  • Rely on automatic webhooks for ongoing data rather than large on-demand pulls.

Last updated

Was this helpful?