Hi,
I've just been doing some prototyping of incorporating a two feature layers (X. poly), one for purpose of editing, one for purpose of snapping (Y. line). Feature layer X is a light-weight poly dataset whereas feature layer Y is a ultimately a heavy dataset (OSMM topo layer (line) for those familiar with Ordnance Survey data (UK)).
I have both an issue (performance) & a couple of queries:
1. Performance for feature layer Y is unacceptably slow to draw (I say this in light of the performance provide via the JS API for similar prototype environment (which is almost instantaneous (1-2 sec) whereas SL performance is the order of minutes (approx 5 mins)). Feature service is set-up for OnDemand mode & i have scale range set to only display below 1:500. Note my OnDemandCacheSize is just default (1000). SL API v2.3 / AGS 10.1 SP2.
What gives @ESRI? Is this a consequence of a poor transport model for data via Silverlight from server to client? Is this an ESRI SL API issue? Are things likely to improve in the future with the SL API?
2. When dealing with v. large feature layers, is there any recommended approaches (other than simplifying your data)? Does the fact i have a scale ranges set at MXD have any impact on the features being retrieved by the service on startup extent? Is there any point in caching the data (noting i just want this feature layer as a reference layer).
Appreciate any insight into this topic.
Brad
