Re: Loading one big table into an even bigger table...

The only problem that I see is the bulk collect pulls all 900,000 into memory at once. Unfortunatly a insert into select from I believe will fill up our rollback segment space...

Has anyone used bulk collect to process a very large number of rows?

"Jim Kennedy" <kennedy-family_at_home.com> wrote in message news:<1HP47.360799$p33.7312816_at_news1.sttls1.wa.home.com>...
> insert into tableBig as select .... from tableSmall where...
> commit;
> That's the fastest way to do it. Much faster than plsql cycling through the
> records and inserting them even in an array fashion (bulk collect.)
> Jim
> "Simon" <stemp1ar_at_yahoo.com> wrote in message
> news:8fe904ce.0107162023.70bc871b_at_posting.google.com...
> > I am very new to 8i and I would like to enhance a stored proc using
> > some of the features in Oracle 8i.
> >
> > Currently I have a cursor that selects all records from one table and
> > inserts them into another table since the two tables reside within the
> > same schema and are identical(one is a table containing 1 month of
> > data while the other contains all months of data). I dont believe I
> > can use export import because primary key is being created using a
> > trigger with a sequence generator.
> >
> > I have looked at the bulk collect and the problem is the smaller table
> > has _at_ 900,000 records and is growing.
> >
> > What ways would you suggest to speed up the process using oracle 8i
> > features?
Received on Wed Jul 18 2001 - 15:07:17 CEST