I have been working on an editing application and I would like to understand (due to performance concerns) how the server/services interact with the browser. The FeatureLayers I am using are in On Demand mode.
In the browser console, Network tab, I am looking at objects are downloaded by my browser, along with other information on size and time. When I am at the viewable scale, there are these XMLHttpRequest objects, about 300-400 bytes long. In one particular application, I have 200 of them. Most of them usually seem identical except for the geometry (xmin, xmax, ymin, ymax). They don't have any features in them. I suppose these are some type of queries, is that the case? Why does the browser receive these?
Then there are the XMLHttpRequest objects with the features' information. All these objects are downloaded by the browser even if I am viewing an area without any of the features (e.g. middle of the ocean). They are small objects (again, about 300-400 bytes long), however, they take sometimes up to a second. Also the editor works slower. For example, the line appears half a second after I click for editing for a line feature, and the attribute editor shows significantly later after I finish the geometry editing by double clicking. Is there a way to optimize browser interaction? Are there best practices?