[Benchmarking] Testing setup

Adrian Custer adrian.custer at geomatys.fr
Mon Aug 9 05:04:22 EDT 2010


Hey all,

I've been working on developing a proposal for a testing regime that
would form a useful basis for the future. My goal is to work from the
WMS and SLD specs themselves taking into account different purposes for
which WMS servers are built and therefore the different possible
designs, to develop a good strategy for discrimination between servers. 

However, that work is progressing slowly, probably too slowly for this
testbed so, in the meantime...


Formats:
        raster only queries => JPEG 
        vector only queries => PNG24+PNG8
        mixed queries => JPEG?


Projections:
        (vector only?) EPSG:22662, Platte-Carré,    (no reproj)
        EPSG:25831, UTM       (original, no reproj for raster)
        EPSG:3857,  google    (reproj, no datum change)
        EPSG:23031, UTM/ED50  (reproj, with datum change)
        Lambert Conformal Conic, 60N,30N; 20East    (not mercator)
        
        The third is a projection which has been commonly used in Spain
        earlier in the 20th century while the last is the projection
        used for the recently completed map of European geology and
        might also have an EPSG code but we have not yet searched the
        database. An alternative Lambert Conformal Conic is EPSG:2062 on
        a Madrid meridian but it uses its own, unique datum so may not
        be widely shared.

Envelopes: min i, min j, max i, max j
        raster
                EPSG:25831
                        372690,4576390,429130,4622010 (= pixel density)
                        372360,4557880,484490,4631460 (full area)
        vector
                EPSG:22662
                        -6 E, 38 N, -1 E, 43 N (= pixel density)
                        -10.0E, 36N, 6E, 44N (full area)
        
        The former envelopes define a zone in which all requests will
        cover an area with information rather than empty zones like the
        sea. The third bbox serves the same purpose as Andrea's proposal
        (either works):

On Thu, 2010-08-05 at 16:20 +0200, Andrea Aime wrote:
> If we limit ourselves to
> - lat between 37.5 and 42.7
> - lon between -6 and -1
> we'll basically get something for all vector requests



These proposals were extracted from the discussion for more compact
reading. Other, less structured, proposals and discussion will come
through other mail.

--adrian





More information about the Benchmarking mailing list