For information, setting the variable DEFAULT_IN_MEMORY_LINESTRING in
PostgreSqlDatasetWriterFactory.java to true caused the import to run much
faster - about 1.5 hours. It may also have helped that I set the geometry
correctly with the new geometry column, i.e:
SELECT AddGeometryColumn('ways', 'linestring', 4326, 'LINESTRING', 2);
Which I hadn't done before.
Regards
Steve
-----------------------------------------------------------------------------------------------------------------------
I'm using Osmosis to import a recent UK extract of the planet file into a
database, and I used the simple schema here:
http://svn.openstreetmap.org/applications/utils/osmosis/script/pgsql_simple_schema_0.6.sql
This took around 1 hour.
I then found that this schema didn't give way geometries so re-ran the query,
adding a geometry column 'linestring' so that way objects would be created.
This import has taken 24 hours so far and isn't yet finished. Is this normal or
has something gone wrong? It would be useful to have some sort of counter to
find out how far into the process osmosis is - for example roughly what line of
the XML it is on, as a percentage of the whole file.
Admittedly I am going against advice and using .bz2 files, but it appears
postgresql is causing all the overhead.
Thanks
Steve
_______________________________________________
dev mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/dev
