Problem statement:

Harness APIs do not support rate limit header values like their Split counterparts did.

Ask:

Enhance Harness APIs to return rate limit headers in the response of each API calls

Sample API endpoint where we would like these headers returned if possible: https://apidocs.harness.io/user/getaggregatedusers

Reasoning:

We implemented a ‘leaky bucket’ approach to rate limiting using the following headers from Split API:

x-ratelimit-remaining-ip

x-ratelimit-reset-seconds-ip

x-ratelimit-remaining-org

x-ratelimit-reset-seconds-org

Without these headers or similar we run into rate limit issues that result in 429 error when jobs overlap. I would prefer to have preemptive guard rails in place like we do now instead of relying on exponential retries or other reactive solutions.