On Wednesday 03 May 2006 16:12, Larry Rosenman wrote: >Javier de la Torre wrote: >> It is inserts. >> >> I create the inserts myself with a Python programmed I hace created >> to migrate MySQL databases to PostgreSQL (by th way if someone wants >> it...) > >Ok, that makes *EACH* insert a transaction, with all the overhead. > >You need to batch the inserts between BEGIN;/COMMIT; pairs, or, better >yet set it up as a COPY.
I'm using essentially the same approach for my custom backup/restore procedure. I also found it a very slow process. But when I wrapped up each table script (ie. 20-30k of INSERTs) the time it took to populate the entire database went down from about half an hour to 50 seconds. Very impressive ;-) However, I'm wondering if there's a practical limit to how many rows you can insert within one transaction? -- Leif Biberg Kristensen :: Registered Linux User #338009 http://solumslekt.org/ :: Cruising with Gentoo/KDE ---------------------------(end of broadcast)--------------------------- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
