We have a central table (stored in SDE within SQL Server) containing ~50,000 points. This should more or less be distributed through Server as 50 separate services containing 1,000 points each. We have a web front end based on the JavaScript API that we use to access the services. There seem to be 3 basic ways to do this:
- 50 Separate Services: Add a query layer to ArcMap selecting the 1,000 points for Service A and then publishing that layer as a stand-alone service. Then repeat for the other 49 layers, each published as a stand-alone service.
- 1 Service with 50 Layers: Create a single service with 50 unique query layers each accessing their 1,000 points and then using the API to access the individual layers of the single service as needed.
- 1 Service with 1 Layer: Create a single service with 1 query layer accessing the entire original table and then using QueryTasks within the API to separate the data at that point.
Acknowledging that every implementation is unique, are there any rules of thumb as to performance, resource allocation, etc. that would make any one method better than the others?
Thanks.
