[mapserver-users] MapServer Oracle performance

[Smith, Michael ERDC-USACE-NH]

Ivan, Bart, Howard,

I created ticket #3107 to include this patch. Also in this patch is support
for 4d data (x,y,z,m) since that's in my version of maporaclespatial.c as
well.

And yes, this is for Mapserver's internal Oracle support.

Performance is much faster that Oracle's MapViewer in my experience. The
performance increase gets even larger, with larger datasets.

Please feel free to make any code improvements you can. C is not my best
language.

Mike


-- 
Michael Smith
RSGIS Center
US Army Corps of Engineers



On 8/27/09  8:28 PM, "Ivan" <ivan.lucena at pmldnet.com> wrote:

> Hi Mike,
> 
> I am not that familiar with the problem but I believe you are talking about
> the internal MapServer's
> Oracle SDO_GEOMETRY support not GDAL/OGR/OCI. Right? Would that run faster
> with? How does it compare
> in performance with Oracle's Mapviewer?
> 
> Anyway, I believe I understand the issues of Binding in OCI since I faced it
> in GDAL/GeoRaster
> myself. So if you need one more eye on the code, count me in.
> 
> But since we are talking about the overall performance of MapServer with
> Oracle I would like to take
> the opportunity to ask a question that concerns me, if you don't mind.
> 
> I setup an very simple MAP file with a GDAL/GeoRaster as the DATA parameter of
> a layer like that:
> 
>    LAYER
>      DATA "geor:scott/tiger at orcl,CITIES,SCENE,ID=120"
> 
> This raster is 14336x14336x3 big (600Mb) and has 7 level of pyramids on it.
> The question is that,
> because of the pyramids (= GDAL overviews) this image loads, pan and zoom
> pretty fast on QuantumGIS
> with the Oracle Raster plugin but that doesn't help at all in MapServer.
> 
> I might not be that familiar with MapServer architecture either, but it looks
> to me that it doesn't
> take advantage of Pyramids at all. Not even if the raster is stored in MrSID.
> The only way to go is
> to build tile cache. Right?
> 
> Best regards,
> 
> Ivan
> 
> 
> Smith, Michael ERDC-CRREL-NH wrote:
>> All,
>> 
>> This is true and I have a patch ready to address this and use bind variables
>> in maporaclespatial.c. Howard (hobu) has the patch and was going to evaluate
>> it and incorporate into trunk when he gets a chance.
>> 
>> I haven't seen an amazing boost in performance but it does reduce the amount
>> of duplicate sql in the SGA.
>> 
>> Mike
>> 
>> 
>