On Sun, May 22, 2011 at 10:20 AM, yobiSource <[email protected]>wrote:
> Am 15.05.2011 21:33, schrieb Andrea Aime:
> > Wow, that's a lot already, it means the map will consume 23 times more
> > memory than normal. Seems a patologic case already, you sure you
> > really need 23 different z orders in that map?
> > If so the code should probably revert to a multiscan approach, taking
> > 23 times longer instead of 23 times more memory. We don't have
> > this kind of heuristic now.
> >
> > There is no limit on the feature type style per se but:
> > - every time you create a new FTS a new drawing surface is allocated,
> > meaning you march steadily towards a OOM
> > - GS ships with built-in WMS limits that are there exactly to disallow
> > requests that use too much memory, so you'll probably get sooner
> > a case in which GS refuses to answer WMS requests because they
> > are going to allocate too much memory compared to the limit
> >
> > Cheers
> > Andrea
> >
>
> Is it better to create five layers, each with five FeatureTypeStyle?
>
It's a compromise. That will likely use 5 times more time but consume
5 times less memory than the case with 25 feature type styles
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 962313
http://www.geo-solutions.it
http://geo-solutions.blogspot.com/
http://www.youtube.com/user/GeoSolutionsIT
http://www.linkedin.com/in/andreaaime
http://twitter.com/geowolf
-------------------------------------------------------
------------------------------------------------------------------------------
What Every C/C++ and Fortran developer Should Know!
Read this article and learn how Intel has extended the reach of its
next-generation tools to help Windows* and Linux* C/C++ and Fortran
developers boost performance applications - including clusters.
http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________
Geoserver-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-users