[OSM-dev] [Mapnik-users] osm2pqsql caching strategy

[Erik Johansson]

On Thu, Feb 5, 2009 at 9:05 AM, Nop <ekkehart at gmx.de> wrote:
>
> Hello!
>
> Currently I am using osmosis to cut the input data for my map into
> sections. This cutting method drops data along the borders and produces
> ugly "creases" on the map. The mode completeWays=yes is too slow for
> planet files.
>
> So I am thinking about implementing my own data cutter or a seamless
> data conversion process. Due to  memory considerations, both will need a
> database-driven node cache to work, just like osm2pqsql in --slim mode does.
>
> The caching of osm2pqsql seems pretty effective, so I wonder what is the
> chaching stragy of the tool? What nodes are kept in cache? What data is
> written to DB?
>
> Or do you have any other experiences with other tools for node caching?

Openstreetmap question..

You want to import the whole planet, but in sections? Just use
osm2pgsql to import the whole planet and then select sections from the
postgis database, initial import is going to be slower but keeping it
uptodate is going to be easy. Alternatively import a bigger area than
what you need and do the clipping in postgis.