But why then is the speed acceptable if I copy and then manually add the FK? Is the check done by the FK so much different from when it is done automatically using an active deffered FK?
Yeah I think it uses a different query formulation... Actually I only assume that deferred fk's don't use that - I guess your experiment proves that.
In my tests deferred or not deferred makes no difference in speed. I am still quite surprised by how huge the difference is.. this makes FKs quite unusable when added a lot of data to a table.
Actually, you can just "disable" them if you want to be really dirty :)
Thanks for the pointer. I got this from the archives:
------------------------ update pg_class set reltriggers=0 where relname = 'YOUR_TABLE_NAME';
to enable them after you are done, do
update pg_class set reltriggers = count(*) from pg_trigger where pg_class.oid=tgrelid and relname='YOUR_TABLE_NAME';
------------------------
I assume the re-enabling will cause an error when the copy/insert added data that does not satisfy the FK. In that case I'll indeed end up with invalid data, but at least I will know about it.
Thanks,
-- Richard van den Berg, CISSP ------------------------------------------- Trust Factory B.V. | www.dna-portal.net Bazarstraat 44a | www.trust-factory.com 2518AK The Hague | Phone: +31 70 3620684 The Netherlands | Fax : +31 70 3603009 -------------------------------------------
---------------------------(end of broadcast)--------------------------- TIP 7: don't forget to increase your free space map settings
