Use http rate limiting headers to decide when to retry requests

[aaguiarz]

Checklist

• I agree to the terms within the OpenFGA Code of Conduct.

Describe the problem you'd like to have solved

We are currently retrying requests on 429s based on a hard coded timestamp.

Describe the ideal solution

We should use the http headers to retry. In the example below we'd wait 6 seconds:

Retry-After: 6

Alternatives and current workarounds

No response

References

No response

Additional context

No response

[aaguiarz]

An example of a rate limited request

HTTP/1.1 429 Too Many Requests
Date: Wed, 08 Jan 2025 18:50:01 GMT
Content-Type: application/json
Content-Length: 63
Connection: close
Server: nginx
Retry-After: 6
Strict-Transport-Security: max-age=63072000; includeSubDomains; preload
]X-Content-Type-Options: nosniff
X-Ratelimit-Limit: 10
X-Ratelimit-Remaining: 0
X-Ratelimit-Reset: 1736362261

{
  "code": "rate_limit_exceeded",
  "message": "rate limit exceeded"
}