Hi Jeroen,
I have prepared a pull request containing your updates as well as my fix.
While waiting for the merge, do you have any chance to test if the updated
code solves your issue? (The test you added, is currently passing)
https://github.com/geosolutions-it/geoserver/tree/GEOS-6564
Cheers,
Daniele
==
GeoServer Professional Services from the experts! Visit
http://goo.gl/NWWaa2 for more information.
==
Ing. Daniele Romagnoli
Senior Software Engineer
GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
http://www.geo-solutions.it
http://twitter.com/geosolutions_it
-------------------------------------------------------
On Wed, Jul 9, 2014 at 4:13 PM, Daniele Romagnoli <
[email protected]> wrote:
> Ok. I have found the reason.
> The NetCDF output format support has been developed with a set of current
> working assumptions (which may be relaxed with some additional development):
> Quoting only one of them, from
> http://docs.geoserver.org/2.4.1/user/community/netcdf-out/index.html
>
> - Input coverages/slices should share the same bounding box (lon/lat
> coordinates are the same for the whole ND cube) [That is a typical NetCDF
> use case]
>
> So that the stacked granules are requested to the common bbox.
> In case a set of granules share the same dimension values (but bbox), we
> can consider merging them into a single coverage with the "super" bbox. I
> should think about it a bit (as well as corner cases).
>
> Daniele
>
>
> ==
> GeoServer Professional Services from the experts! Visit
> http://goo.gl/NWWaa2 for more information.
> ==
>
> Ing. Daniele Romagnoli
> Senior Software Engineer
>
> GeoSolutions S.A.S.
> Via Poggio alle Viti 1187
> 55054 Massarosa (LU)
> Italy
> phone: +39 0584 962313
> fax: +39 0584 1660272
>
> http://www.geo-solutions.it
> http://twitter.com/geosolutions_it
>
> -------------------------------------------------------
>
>
> On Wed, Jul 9, 2014 at 3:40 PM, Daniele Romagnoli <
> [email protected]> wrote:
>
>>
>>
>> ==
>> GeoServer Professional Services from the experts! Visit
>> http://goo.gl/NWWaa2 for more information.
>> ==
>>
>> Ing. Daniele Romagnoli
>> Senior Software Engineer
>>
>> GeoSolutions S.A.S.
>> Via Poggio alle Viti 1187
>> 55054 Massarosa (LU)
>> Italy
>> phone: +39 0584 962313
>> fax: +39 0584 1660272
>>
>> http://www.geo-solutions.it
>> http://twitter.com/geosolutions_it
>>
>> -------------------------------------------------------
>>
>>
>> On Wed, Jul 9, 2014 at 3:41 PM, Jeroen Dries <[email protected]>
>> wrote:
>>
>>> Hi Daniele,
>>>
>>> the test does show the issue: have a look at the bounding boxes of the
>>> returned coverages: they do not correspond to the boxes of the original
>>> granules, instead both have a bounding box equal to the one which is
>>> requested. All of the other dimensions are also the same, so in fact the
>>> returned coverages are equal.
>>>
>>> I would expect that either:
>>> -> I get a single coverage with bounding box equal to the request
>>> bounding box
>>> OR
>>> -> I get two coverages, containing the intersection of my request
>>> bounding
>>> box, with the bounding box of the underlying granule.
>>>
>> Ah ok... I was thinking that the test was returning 2 coverages with the
>> intersected bbox... I didn't notice that they share the same bbox.
>> Sorry for that. I'm going to check it again.
>>
>>
>>>
>>> I prefer the first option, because in the second case I would have to
>>> combine
>>> the two coverages into a single one in my output format. What do you
>>> think?
>>>
>>> best regards,
>>> Jeroen
>>>
>>>
>>>
>>> On Wednesday 09 July 2014 15:26:45 Daniele Romagnoli wrote:
>>> > Hi Jeroen,
>>> > I have checked your pull request and your tests but I didn't understand
>>> > where the problem is.
>>> > You are configuring a mosaic made of 2 granules
>>> >
>>> > 10,70 -------- 30,70 35,70 -------- 55,70
>>> >
>>> > | | *&* | |
>>> >
>>> > 10,40 -------- 30,40 35,40 -------- 55,40
>>> >
>>> >
>>> > Then, you are requesting the full bbox so you are getting back both of
>>> them.
>>> > Not sure why you should get back just one of them, since they are 2
>>> > different files/granules composing the result.
>>> > Are you really sure that the test you added on the pull request is
>>> > replicating the issue you are referring to?
>>> > If affirmative, I think that this is not an issue. but it's the
>>> expected
>>> > behaviour.
>>> >
>>> > Please, Let me know.
>>> > Regards,
>>> > Daniele
>>> >
>>> > .
>>> >
>>> >
>>> >
>>> >
>>> >
>>> > So
>>> >
>>> >
>>> >
>>> > ==
>>> > GeoServer Professional Services from the experts! Visit
>>> > http://goo.gl/NWWaa2 for more information.
>>> > ==
>>> >
>>> > Ing. Daniele Romagnoli
>>> > Senior Software Engineer
>>> >
>>> > GeoSolutions S.A.S.
>>> > Via Poggio alle Viti 1187
>>> > 55054 Massarosa (LU)
>>> > Italy
>>> > phone: +39 0584 962313
>>> > fax: +39 0584 1660272
>>> >
>>> > http://www.geo-solutions.it
>>> > http://twitter.com/geosolutions_it
>>> >
>>> > -------------------------------------------------------
>>> >
>>> >
>>> > On Mon, Jul 7, 2014 at 10:50 AM, Daniele Romagnoli <
>>> >
>>> > [email protected]> wrote:
>>> > > Thanks Jeroen,
>>> > > I have seen that JIRA.
>>> > > I'll take a look on this as soon as I can.
>>> > >
>>> > > Thanks again for having set it up
>>> > > Cheers,
>>> > > Daniele.
>>> > >
>>> > > ==
>>> > > GeoServer Professional Services from the experts! Visit
>>> > > http://goo.gl/NWWaa2 for more information.
>>> > > ==
>>> > >
>>> > > Ing. Daniele Romagnoli
>>> > > Senior Software Engineer
>>> > >
>>> > > GeoSolutions S.A.S.
>>> > > Via Poggio alle Viti 1187
>>> > > 55054 Massarosa (LU)
>>> > > Italy
>>> > > phone: +39 0584 962313
>>> > > fax: +39 0584 1660272
>>> > >
>>> > > http://www.geo-solutions.it
>>> > > http://twitter.com/geosolutions_it
>>> > >
>>> > > -------------------------------------------------------
>>> > >
>>> > >
>>> > > On Mon, Jul 7, 2014 at 10:37 AM, Jeroen Dries <[email protected]>
>>> > >
>>> > > wrote:
>>> > >> Hi Daniele,
>>> > >>
>>> > >> I created a JIRA issue:
>>> > >>
>>> > >> http://jira.codehaus.org/browse/GEOS-6564
>>> > >>
>>> > >>
>>> > >>
>>> > >> I also created a unit test, which mimicks my setup, and exposes the
>>> > >> issue:
>>> > >>
>>> > >> https://github.com/geoserver/geoserver/pull/639
>>> > >>
>>> > >>
>>> > >>
>>> > >> It took my a bit of time, but I hope I can reuse this unit test
>>> setup in
>>> > >> the future to illustrate other issues. In fact it also a nice
>>> > >> demonstration
>>> > >> of a complex multi dimensional setup.
>>> > >>
>>> > >>
>>> > >>
>>> > >> best regards,
>>> > >>
>>> > >> Jeroen
>>> > >>
>>> > >> On Wednesday 02 July 2014 16:42:57 Daniele Romagnoli wrote:
>>> > >> > Hi Jeroen,
>>> > >> >
>>> > >> > If you also have a small sample mosaic (configuration and data)
>>> and a
>>> > >> >
>>> > >> > request to replicate it, to be attached to JIRA, they would be
>>> very
>>> > >>
>>> > >> helpful
>>> > >>
>>> > >> > for debugging and fixing it.
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > Cheers,
>>> > >> >
>>> > >> > Daniele
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > ==
>>> > >> >
>>> > >> > GeoServer Professional Services from the experts! Visit
>>> > >> >
>>> > >> > http://goo.gl/NWWaa2 for more information.
>>> > >> >
>>> > >> > ==
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > Ing. Daniele Romagnoli
>>> > >> >
>>> > >> > Senior Software Engineer
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > GeoSolutions S.A.S.
>>> > >> >
>>> > >> > Via Poggio alle Viti 1187
>>> > >> >
>>> > >> > 55054 Massarosa (LU)
>>> > >> >
>>> > >> > Italy
>>> > >> >
>>> > >> > phone: +39 0584 962313
>>> > >> >
>>> > >> > fax: +39 0584 1660272
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > http://www.geo-solutions.it
>>> > >> >
>>> > >> > http://twitter.com/geosolutions_it
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > -------------------------------------------------------
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > On Wed, Jul 2, 2014 at 4:38 PM, Jeroen Dries <
>>> [email protected]>
>>> > >>
>>> > >> wrote:
>>> > >> > > Hi all,
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > I'm seeing some strange behaviour in WCSDimensionsSubsetHelper
>>> in
>>> > >> > >
>>> > >> > > combination
>>> > >> > >
>>> > >> > > with a multi-dimensional mosaic:
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > ->In the splitRequest method, all of the granules are queried
>>> for a
>>> > >>
>>> > >> given
>>> > >>
>>> > >> > > bounding box, and set of dimensions. In the case of a mosaic, a
>>> > >>
>>> > >> granule
>>> > >>
>>> > >> > > exists
>>> > >> > >
>>> > >> > > for each tile in the mosaic, and for each dimension.
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > So for example if my bounding box covers 2 tiles, and my
>>> dimension
>>> > >>
>>> > >> range
>>> > >>
>>> > >> > > also
>>> > >> > >
>>> > >> > > intersects with 2 discrete values, I get 4 granules in total.
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > -> Next the weird thing happens, for all of the 4 granules, a
>>> new
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > subrequest
>>> > >> > >
>>> > >> > > is made. The bounding box remains the original bounding box,
>>> but the
>>> > >>
>>> > >> value
>>> > >>
>>> > >> > > for
>>> > >> > >
>>> > >> > > the dimension is correctly extracted from the granule.
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > The result is that in the final output, I now have 2 identical
>>> pairs
>>> > >>
>>> > >> of
>>> > >>
>>> > >> > > coverages, which seems wrong.
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > I hope that this description is a bit clear, if no one objects,
>>> I'll
>>> > >>
>>> > >> log
>>> > >>
>>> > >> > > this
>>> > >> > >
>>> > >> > > in Jira.
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > Not sure what the right solution is, perhaps the subRequests
>>> can be
>>> > >>
>>> > >> stored
>>> > >>
>>> > >> > > in
>>> > >> > >
>>> > >> > > a Set, so that only unique requests are retained...
>>> > >> > >
>>> > >> > >
>>> > >> > >
>>> > >> > > best regards,
>>> > >> > >
>>> > >> > > Jeroen
>>> > >>
>>> > >>
>>> -------------------------------------------------------------------------
>>> > >> -
>>> > >>
>>> > >> > > ---- Open source business process management suite built on
>>> Java and
>>> > >> > >
>>> > >> > > Eclipse Turn processes into business applications with Bonita
>>> BPM
>>> > >> > >
>>> > >> > > Community Edition Quickly connect people, data, and systems into
>>> > >> > >
>>> > >> > > organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner
>>> awards
>>> > >> > >
>>> > >> > > http://p.sf.net/sfu/Bonitasoft
>>> > >> > >
>>> > >> > > _______________________________________________
>>> > >> > >
>>> > >> > > GeoTools-Devel mailing list
>>> > >> > >
>>> > >> > > [email protected]
>>> > >> > >
>>> > >> > > https://lists.sourceforge.net/lists/listinfo/geotools-devel
>>>
>>>
>>
>
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
GeoTools-Devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel
