submitting a pull request for an initial level of integration...
I made sure that everything is wired as it was, to wire in geowave support, modify line 175 of TileCacheHibernate.groovy to swap bean definitions
This initial cut is meant to be 0 impact...the reality is you will get nothing from this the way it is currently being used (what you may get is the ability to use GeoServer for a different usage) - it just introduces slight inefficiency for the current usage...
because PostGIS is relied on for all the indexing, and the current spatial query is tied to PostGIS to bring back a series of keys, the "0 impact" way I tied in GeoWave was to take the bounds from PostGIS results and use it to grab the tiles - kinda counter-intuitive, but to do more would probably impact the current query logic
Because GeoWave is the spatial index on the raster data, the way it is supposed to be used is to write data with a spatial extent and query the data with a spatial extent.
this initial cut will query postGIS, bring back results with the bounds and use each tile's bounds to then get the data. In particular when a user wants to query spatially, postGIS would typically be unnecessary with GeoWave - just go straight to geowave with the query geometry
submitting a pull request for an initial level of integration...
I made sure that everything is wired as it was, to wire in geowave support, modify line 175 of TileCacheHibernate.groovy to swap bean definitions
This initial cut is meant to be 0 impact...the reality is you will get nothing from this the way it is currently being used (what you may get is the ability to use GeoServer for a different usage) - it just introduces slight inefficiency for the current usage...
because PostGIS is relied on for all the indexing, and the current spatial query is tied to PostGIS to bring back a series of keys, the "0 impact" way I tied in GeoWave was to take the bounds from PostGIS results and use it to grab the tiles - kinda counter-intuitive, but to do more would probably impact the current query logic
Because GeoWave is the spatial index on the raster data, the way it is supposed to be used is to write data with a spatial extent and query the data with a spatial extent.
this initial cut will query postGIS, bring back results with the bounds and use each tile's bounds to then get the data. In particular when a user wants to query spatially, postGIS would typically be unnecessary with GeoWave - just go straight to geowave with the query geometry
...so yeah, you could put the data in geowave, add geoserver, and try out geotools' build-in support for geopackage (http://docs.geoserver.org/latest/en/user/community/geopkg/output.html) ...or not...