Justin Deoliveira ha scritto:
> Hi Jeffrey,
> 
> You are right in that you need to get rid of MemoryDataStore. It sounds 
> like the reason you are using it is to be able to work with multiple 
> files at once. If that is the case, you may wish to consider loading all 
> of your shapefiles into a postgis database, and using a postgis 
> datastore instead.
> 
> Serving your files via wfs with GeoServer is an option too, but I think 
> the postgis solution is simpler.
> 
> Also, as I am sure you are aware the graph module builds entire graphs 
> in memory which leads to a big memory requirement. What I have done in 
> the past to get around this is not store geometries or features in the 
> graph. For instance for a line network, Just storing an x,y coordinate 
> at each node, and an identifier at each edge. Then when an actual edge 
> attribute or geometry is needed it is loaded on demand from the database.

I'm wondering if a cache design like the one described in the Datastore 
caching R&D page 
(http://docs.codehaus.org/display/GEOTOOLS/Datastore+caching) could
be useful in this case... it seems so, I mean, the page assumes caching
feature is what you want, but the general caching approach described
at the bottom of the page could be used for anything that has a spatial
component, so maybe it could apply to graph links too?

The idea would be to keep a virtual graph in memory, and keep only the
part you're visiting in memory at a time (with caching of course to
avoid frequent and useless reloading).

Cheers
Andrea

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Geotools-gt2-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users

Reply via email to