Access token quota

Question

The documentation says that each user gets 5 distinct quotas of 10,000 requests per 24 hours, and as I interpret it, per app.

If an application does have an access_token, then the application is on a distinct user/app pair daily quota (default size of 10,000). A user can have up to 5 distinct quotas at any one time, though this limit is not reflected in quota_remaining returns for privacy reasons.

How is the quota managed if the user has more than 5 apps using access tokens? will 2 of the apps share one 10,000 block? or can the user only use 5 apps (that use access tokens) at once?

Basically should I expect the possibility of having no quota to use, because the user already has 5 apps?

Answer 1

Each application/user pair (as signified by an access_token) has a per-day quota of 10k requests (by default). Any number of these can be in flight at once, and they're reported as part of quota_remaining and quota_max.

On top of that, is a per-user quota of 50,000. This quota isn't reported anywhere (for privacy reasons, we'd rather not leak a user's request behavior between apps) but when it's exhausted we deny all requests from that user. You'll get a slightly different error message, though the code will still be throttle_violation.

In short, there isn't a "5 active apps" limit; but you do need to be able the case where the user are themselves throttled. There's not much you can do beyond through up an error message in that case though. It should be quite rare, given that a typical application shouldn't use anywhere near it's full 10k quota nor should a typical user have more than 5 apps (if even that) active at once.