[FeatureServer] fast drawing of large data amounts

Iocast iocast at me.com
Sat Apr 13 02:17:50 PDT 2013


Hi Paul,

We are using FeatureServer to serve the OSM data for switzerland. The data are stored in a PostgreSQL database.

A) Using OGC's filter encoding definition language you are able to reduce the amount of data that will be send back to the client.

B) In addition we use a so called FeatureFilter server that does a server-side clustering. The workflow is as follow
1. Client Request to FeatureFilter --> 2. forward request to WFS server --> 3. get response from WFS server --> 4. cluster it --> 5. send back result to client

C) Currently this clustering server is a prototype and the performance need to be improved. You are welcome to enhance it (source: https://github.com/iocast/featurefilter)

hopefully I could help you
Michel



On Apr 12, 2013, at 7:12 PM, Paul Brossier <piem at piem.org> wrote:

> Hi,
> 
> I am trying to load and draw my travels routes using featureserver.
> 
> The original data is a mix of large kml files and daily gpx log files. I
> cleaned it up, converted it to geojson and loaded everything in the server.
> 
> But it is still a huge amount of data (about 11M in dbm files).
> 
> What are the best practice to efficiently serve so much data, being kind
> on both the server and the client sides?
> 
> Ideally, I guess I would need to implement a Douglas-Peucker algorithm
> on the server side to serve less points when the zoom level is small.
> 
> Thanks, Paul
> _______________________________________________
> FeatureServer mailing list
> FeatureServer at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/featureserver



More information about the FeatureServer mailing list