It's probably worth reporting if you can find a test case. However, I'd first suggest:
- Ruling out caching effects. It's not just because the 8.3 database has been running for a while already and the 8.4 has just started up? - Are the two servers configured the same. Compare the config files, is the same tuning applied to both? At the psql promt you can do a "SHOW ALL" to see all the parameters. A quick diff will tell you about major changes. Hope this helps, On Wed, May 12, 2010 at 2:09 AM, Pieren <pier...@gmail.com> wrote: > On Sat, May 8, 2010 at 1:11 AM, Frederik Ramm <frede...@remote.org> wrote: >> >> The intial import went all right on 8.3 and 8.4; my "4 times slower" was >> about adding a daily diff to the data base. >> > > Just for information, I also compared the performance on my Ubuntu 10.04 > Lucid, amd64 distrib (ext4, hdd raid0 10k rpm) and I got the following > numbers: > My country new import: > 8.4: 33mn07 > 8.3: 32mn46 > > Daily diff import: > 8.4: 43mn40 > 8.3: 26mn09 > I don't have your factor 4 but 1.6x slower with postgresql 8.4. > > Pieren > > _______________________________________________ > dev mailing list > dev@openstreetmap.org > http://lists.openstreetmap.org/listinfo/dev > > -- Martijn van Oosterhout <klep...@gmail.com> http://svana.org/kleptog/ _______________________________________________ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
