On Thu, Feb 5, 2009 at 9:05 AM, Nop <[email protected]> wrote: > 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?
Check out the comment in middle-pgsql.c, it describes how it works. *Everything* gets written to the database, just there's also a cache in memory attempting to remember the maximum number of nodes for a given amount of memory. Nodes are added to blocks as they come and a priority queue remembers which block has the least used entries. When a new block is needed it takes the most empty one first. It requires for maximum efficiency that the nodes come in ID order. Have a nice day, -- Martijn van Oosterhout <[email protected]> http://svana.org/kleptog/ _______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
