I published a Postgis-raster table as a Mapserver WCS layer using
gdal-1.10dev driver. I'm getting the expected result as output of my
Getcoverage query but I'm concerned by the time required to extract the
data. It takes about 3 minutes to extract a 4800 x 4800 pixels area. I
looked at the SQL queries performed against the database during data
processing and found that the GDAL driver execute a query in order to get
the extent of my layer using the following SQL query:
select srid, nbband, st_xmin(geom) as xmin, st_xmax(geom) as xmax,
st_ymin(geom) as ymin, st_ymax(geom) as ymax
from (select st_srid(rast) srid, st_extent(rast::geometry) geom,
max(ST_NumBands(rast)) nbband from mib_postgis.gridcoverage_
cdsm75 group by st_srid(rast)) foo;
srid | nbband | xmin | xmax | ymin
| ymax
------+--------+-------------------+-------------------+------------------+------------------
4617 | 1 | -139.500104166667 | -52.0001041666667 | 41.5001041666667 |
60.0001041666667
(1 row)
Time: 153033.599 ms
I'm wondering how I could avoid this SQL to be performed on each call,
since it is consuming a large amount of the request processing time and
values are always the same.
Is it something that could be fixed at database level or into my mapfile.
See the my current mapfile layer definition:
LAYER
NAME cdsm
GROUP cdsm
METADATA
"wcs_label" "cdsm" ### required
"wcs_rangeset_name" "Range 1" ### required to support
DescribeCoverage request
"wcs_rangeset_label" "My Label" ### required to support
DescribeCoverage request
"wcs_extent" "-139.500104166667 41.5001041666667 -52.0001041666667
60.0001041666667"
"wcs_resolution" "0.000208333333333 0.000208333333333"
END
TYPE RASTER
STATUS ON
PROCESSING "NODATA=255"
DATA "PG:host=hostname port=portnumber dbname='dbname' user='username'
password='passwd' schema='schema' table='gridcoverage_cdsm75' mode='2'"
PROJECTION
"init=epsg:4617"
END
END
Thanks,
J-F
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev
