Closed si-the-pie closed 10 years ago
Looks more like a postgres configuration problem as the high IO load remains after osm2pgsql has finished. Could you confirm that the settings for work_mem
is at least at 100MB?
Also could check that the number of sequential scans is in the same order as on poldi. (Should not increase during the update).
I've increased the work_mem to 131MB as you suggest, and made several other changes, eg. to maint_work_mem too and performance seems to be much better now, thanks.
These munin charts relate to my question at: https://lists.openstreetmap.org/pipermail/geocoding/2013-November/001037.html
I've added a few of the munin charts that hopefully illustrate the problem.