You are DEFINITELY not going to win any races with INSERT REPLACE.
Minimizing the number of records you need to process will improve your
times significantly, especially since you need less than .002% of your
source data added to your destination table.
Unimin Corporation - Spruce Pine
matt ryan <matt_lists@stripped> wrote on 08/19/2004 12:42:10 PM:
> Andrew Pattison wrote:
> > I'm guessing that you have indexes on the 321st_stat table? If this is
> > the case, try dropping them before you do the insert, then rebuilding
> > them. MySQL is known to be slow at doing bulk inserts on indexed
> > tables. Also, updates are much faster than inserts since with inserts
> > there are much more disk IOs required.
> > Cheers
> Table has a large primary key, to keep duplicates out, so I cant drop
> the index and remove.
> Also, reindexing the table takes all day, dropping is not an option
> Would I be better off doing an insert replace on all 17m new records,
> or only inserting the 500 new records using insert ignore ?
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe: http://lists.mysql.com/mysql?unsub=1