> They PGSQL-PERFORMANCE list is really the appropriate place for
> performance issues like yours. Subscribe?
Josh,
Brian and I are trying to put upwards of 80-90,000 rows into a table.
When we run on PostgreSQL 7.4, this takes about five minutes. On
comparable hardware, running PostgreSQL 8.0.1, it's taking almost two
hours. The query from the bug report shows that the planner is using
sequential searches, not index searches, to satisfy the foreign key
constraint.
Our scenario (inserting data into a new table) seems to be something
that a lot of users could run across. The performance drop we're seeing
shows up as a regression from earlier behaviour in PostgreSQL. That's
why, after discussing our issue with AndrewSN on #postgresql, Brian
filed a bug here.
We'll follow up with the PERFORMANCE list to find a workaround, but I
think this is an issue which should be addressed as a bug.
Thanks for your help,
Keith Browne
tuxedo@deepsky.com