On Thu, Dec 27, 2012 at 12:32 PM, Jonathan Moules <
[email protected]> wrote:
> Definitely have a Spatial Index. I'd know it if I didn't. ;-) (but have
> just double checked - its there).
>
> This issue seems to be a little more nuanced than I previously thought.
>
> Now when I click "Compute from data" Geoserver web interface doesn't
> appear to respond at all. Not even the error. I get a little flower-loading
> icon in the top right and that's it. After a minute or two the new bounding
> co-ordinates are loaded, but for all intents and purposes it actually
> appears that GeoServer has crashed during that period.
>
This is because the inner configuration of GeoServer is not yet thread
safe, the system is old and was created when only
one admin could operate on it at a time.
Unfortunately for long bbox computations the net result is that no other
configuratione page can be opened... but the
OGC services are still operating.
>
>
> I ran the same query in Oracle SQL Developer and it took 2 minutes! That's
> a very long time to get an MBR for just 50,000 features. I dread to think
> what'll happen when I point it at my ~15million feature table!
>
> I think GeoServer should use the MDSYS.USER_SDO_GEOM_METADATA entry by
> default - that's what its there for after all.
>
Of all the spatial databases Oracle has the worst situation when it comes
to proper configuration for spatial tables, I have
yet to see in person a single installation that has all the metadata in
order, be either bbox, or declaring the target geometry
type in spatial indexes, that's probably why no one bothered to implement
the code to read from the metadata in
the last hum... six or seven years.
I believe it once existed, but it caused so much trouble due to
mis-configured instances being the norm that when the
JDBC stores got rewritten a few years ago we did not bother to put it back.
That code could be brought back of course, and could be activated with a
flag (not by default by any means though,
only by those few ones knowing their instance is properly setup).
The Java class to be modified is this one:
https://github.com/geotools/geotools/blob/master/modules/plugin/jdbc/jdbc-oracle/src/main/java/org/geotools/data/oracle/OracleDialect.java
It should be similar to PostGIS "estimated extent" machinery here:
https://github.com/geotools/geotools/blob/master/modules/plugin/jdbc/jdbc-postgis/src/main/java/org/geotools/data/postgis/PostGISDialect.java#L251
Only, in this case, the flag to activate that behavior should probably be
called "declared extents".
Patches welcomed :-)
Cheers
Andrea
--
==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more
information.
==
Ing. Andrea Aime
@geowolf
Technical Lead
GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549
http://www.geo-solutions.it
http://twitter.com/geosolutions_it
-------------------------------------------------------
------------------------------------------------------------------------------
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. ON SALE this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122712
_______________________________________________
Geoserver-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-users
