In the last episode (Aug 19), matt ryan said:
> >One alternative, since you know you don't have many records to insert,
> >is to pull the IDs of the missing records and insert just those. Do an
> >outer join on the two tables (joining on the primary key), get a list
> >of the IDs of records in stat_in but not in 321st_stat, and add a
> >"WHERE id IN (list,of,ids)" clause to the end of your INSERT ... SELECT
> >statement. If you're running 4.1, you can use a subquery and embed the
> >first query directly in the INSERT.
>
> Running 4.0.x
>
> something like, select a.*, b.* from a left outer join b on
> a.col1=b.col1, a.col2=b.col2, a.col3=b.col3 where b.col1 is null into
> temptable
>
> then insert from temptable into table a
>
> I think I tried this once, but it ran all day
The select, or the insert? If the select, run an EXPLAIN and see if
it's using the wrong index or no index.
> Is there a way to join on an index, instead of on each column? The
> primary key is 6-8 columns I forget
You have to list each column. "SHOW KEYS FROM table" or "SHOW CREATE
TABLE table" will list all the keys and which fields are used in each
key.
--
Dan Nelson
[EMAIL PROTECTED]
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
