Chris Koutras wrote:
Wim,
Thanks for the info, particularly the config file values. It is always
good to see what works for someone on a large network. Are you using
the "sql_dont_try_update" directive? I have been thinking of trying
this to reduce the database bottle neck.
I didn't even know that directive existed - must be a new one. So I'm
using the default, which I think is to try an update first.
I suppose it's a toss between extra database operations (try update,
then insert if no records updated) and extra disk space used.
My biggest challenge at the moment is that in order to keep query SELECT
times responsive on tables with 3+ million records, frequent VACUUM
FULLs are required.... but that's very disk intensive and causes the
pmacctd cache purges to backup until the vacuum is done.
(Without vacuuming, the space used by the tables & indexes balloons)
If I have some free time, I may compare PostgreSQL 8.0's performance
against the latest 7.4.
Wim
