[gdal-dev] Recommended OGR Driver For Very Large Vector Data
Jeff McKenna
jmckenna at gatewaygeomatics.com
Fri Aug 3 12:16:42 PDT 2018
My next choice for customers (in cases of a hard file requirement) is
always Spatialite (sqlite3 + libspatialite wrapped into a single file).
I still prefer and use that to deliver, over the heavier GPK format.
(again, everyone has their own preferences, I'm just sharing my 18+
years of MapServer experience with you here ha).
Have a nice weekend,
-jeff
On 2018-08-03 4:08 PM, Patrick Young wrote:
> Ah right of course! I should have said that we deliver these things to
> customers, so a file format is needed. We've have had good luck serving
> simplified versions of these things out of zipped shapefiles from S3 via
> mapserver so the thought was maybe the same trick would work with the
> right format without having to import them into the DB.
>
> P
>
> On Fri, Aug 3, 2018 at 12:52 PM, Jeff McKenna
> <jmckenna at gatewaygeomatics.com <mailto:jmckenna at gatewaygeomatics.com>>
> wrote:
>
> Hi Patrick,
>
> In terms of speed and MapServer, considering you have large vector
> data and require good indexing, I always use and recommend PostGIS.
>
> (of course if you asked this on the MapServer-users list you would
> get a whole lot of different opinions)
>
> But that is my initial recommendation, for your stated needs.
>
> -jeff
>
>
>
> --
> Jeff McKenna
> MapServer Consulting and Training Services
> https://gatewaygeomatics.com/
>
>
>
>
> On 2018-08-03 3:45 PM, Patrick Young wrote:
>
> Hi all,
>
> I've been trying to figure out a good format to store some very
> large vector data in. Basically, its a bunch of really dense
> polygons describing the cutlines that compose a mosaic. We've
> had reasonable success using zipped shapefiles, but on the
> larger guys we're exceeding the 2GB file size, so a more modern
> alternative is desired.
>
> FileGBD worked alright but having to build GDAL with the ESRI
> driver to write them is not desirable. We also tried a
> GeoPackage, but it seemed a little slow to read in QGIS.
>
> The dream would be a nicely compressed format that can be
> spatially indexed so that we can serve them up efficiently with
> mapserver. Any recommendations are greatly appreciated!
>
> Thanks,
> Patrick
>
>
More information about the gdal-dev
mailing list