[OSM-dev] Planet OSM 'bad result during COPY'

[Seth Voltz]

Well, the misconfigured RAM cap was the last missing piece, osm2pgsql seems to have all the resources it requires. The last question I have is one of time. As of an hour ago I've been running the import for 5 days solid. Around 8 hours ago I thought the run was complete as it spat out the last "Completed planet_osm_*" line that I expected to see from a posted example [1]. Pulling up top, I'm still using about 90% of the RAM used during the active import stages and postmaster is using between 3% and 7% of my CPU with occasional bursts to 10% (during the import it ran between 10% and 50% with short spikes higher). I can see the OSM data in the database and query it (though slowly). Is this normal? Given the time it took to import everything, does anyone have an idea how long I can expect to wait before osm2pgsql completes its execution? What is it doing?

Also, is it safe at this point to start up tile generation?

[1] http://www.adamestrada.com/2010/03/30/openstreetmap-planet-osm-processing-statistics/

Thanks,
Seth

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20100419/c1a39355/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 488 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20100419/c1a39355/attachment.pgp>