In a recent APiGEE presentation, the presenter mentioned briefly about sideloading pattern to reduce chattiness of API. I could not find much details on such pattern (when to use, how is it different from use of simple query parameter etc.). Perhaps community can throw more light on the same
Do you mind sharing the link to the presentation you are referring to?
That was me. Side loading is a strategy to avoid repeating elements in a payload. Here is an example:
In this case we only load the product definitions once - at the tail of the payload such that orders may simply refer to them by identity rather than repeating these elements in the lines arrays. This is a trivial example, but in real world scenarios this strategy has reduced payload by as much as 80%.
From a chattiness perspective the products array resolves what otherwise might be multiple individual resource calls to retrieve product details.
Hi @Dallen
This looks similar to attributes in the XML world.
Do you think only API layer(and consumer) should adhere to this pattern or the enterprise data model should take care of it ?
User | Count |
---|---|
2 | |
1 | |
1 | |
1 | |
1 |