any Episerver Find insiders perhaps know? I guess it could be the remote server rejecting requests when made from the same IP address?
Requests will always come from the same IP in a one-server scenario. If you've introduced load testing, could it be you're actually issuing more than 75rps against Find?
Are you caching the result from Find? See http://bergdaniel.se/episerver-find-getresult-getcontentresult for more information.
Keep in mind that one request to a page does not necessary equal one request to the EPiServer Find index. If you have multiple queries on your page, each request to the page will execute each of the Find queries. For example, if you have 5 queries when a page is loaded, 10 page requests will generate 50 requests to the Find index. This will only be the case if you do not cache your queries though, which you really, really, really should.
In the case anyone else hits the limit try to use the multisearch api of Epifind
we've started load testing.
When we send 10 requests per second, EpiServer Find returns with the error: The remote server returned an error: (429) Too Many Requests.
Our production index should be able to handle up to 75 queries per second (and the developer indexes are 25 queries per second).
Is this because we are hitting the index from the same IP address or could it be a configuration issue on our side?