For optimal performance, we have a few recommendations.
Limit query, improve performance
Limiting the scope of the query will directly improve the latency of the API. This can be accomplished by using parameters such as
parameters[date], or (preferably) by using
parameters[updated] for deltas.
A common pattern is to use a delta to query for the latest change in a dataset. By reducing the amount of data available in the query, it will produce results with the least amount of latency and limited data an application will have to traverse.
In our case, we recommend tracking and querying by the latest updated timestamp.
One caveat to allow for some latency and second-only timestamps: when querying, set
parameters[updated] to a value 5 seconds earlier than you actually want. A query may look like:
LAGGED_TIMESTAMP is set to be five seconds less than the maximum value of the
updated field in rows already pulled from our API.
Include key in either query or header