ProfileSync can potentially make a large number of requests, large enough to slow down (if not even overwhelm) your target API. This is especially true for large organizations or for organizations where special events or special promotions might result in huge increases in registrations (and the corresponding user profile creation events). Because of that, it is important that you: 1) have some knowledge of the number of transactions ProfileSync will routinely generate on your behalf; and, 2) have reasonable assurance that your target API will be able to handle all that traffic.
For the most part, there's less concern about ProfileSync's ability to handle large loads: that's because the application includes an intelligent retry mechanism. No matter what you throw at it – network latency, platform availability, API congestion – ProfileSync can gracefully recover from a failed API call and then try again. Note that both the number of retries and the frequency of these retries is configurable.
In addition to that, ProfileSync uses an exponential backoff schedule to help balance retry loads. For example, if the first ProfileSync retry takes place 2 seconds after a failed attempt, the second retry might take place 4 seconds after that, the third retry might take place 8 seconds after that, and so on. In this example, the previous retry interval (e.g., 2 seconds) is multiplied by 2 to determine the time when the next retry takes place (2 x 2 = 4 seconds later).