Storing data in flat database tables - use ArcGIS Rest or custom service?

397
1
04-17-2014 01:22 PM
NicholasFloersch
New Contributor III
Hi folks!

We want to store configurable data for our web application in some flat database tables. Write now we will insert/update these from the back-end. But eventually the users will be able to edit the configuration data in these tables, so they will be read/write. In other words eventually this read only configuration data will turn into a basic sort of CMS.

We are divided here between publishing the postgresql table via SDE as a table in a map service, or using custom web services.

It appears however that you cannot simply create an MXD, put just one table in it (no feature classes, or other map data) and publish it as a map service. So in order to use the ArcGIS Rest API to manage my table I would need at least a dummy feature class in the map service. Any time that I need to put a dummy in place to make something work raises a red flag that I am probably not using the best solution.

The other solution is a custom web service. The down side is that using a custom web service basically means splitting up our configuration tables away from the rest of our data which is all managed in the geodatabase. It would be nice to have everything in one place.

In short - how do other people handle this sort of situation with their JS API web apps? Our JS API app uses AngularJS, so it is not primarily Dojo oriented... if that matters.

Thanks for the input!
Nick
0 Kudos
1 Reply
SteveCole
Frequent Contributor
One of the data layers shown on our web app is a feature layer that's created at page load based on data stored in a standalone table in SDE. Although our web app doesn't edit the table (it's stream gage data so it's just constantly refreshed), I included into the MXD for another feature class that is also shown on our web app. I don't feel that it's too kludgey but the app is somewhat basic and doesn't have a laundry list of datasets to display or edit.

Steve
0 Kudos