Re: Inserts and db file sequential reads

"William Robertson" <williamr2019_at_googlemail.com> wrote in message news:6c24a3aa-971b-4aa0-96e7-a26862d8595a_at_y38g2000hsy.googlegroups.com...
> On May 2, 11:52 pm, oracleing..._at_yahoo.com wrote:
>> INSERT /*+ append */ INTO OPERATION_DTLS
>> VALUES
>
> INSERT /*+ APPEND */ is only applicable to INSERT SELECT, not INSERT
> VALUES. Just as well, as each INSERT /*+ APPEND */ uses new blocks
> above the high water mark rather than looking for space in existing
> blocks, so unless each row uses exactly one block it could waste a
> load of space.

William,

I nearly made the same comment in my post, and then I decided to hedge my bets a little because I wasn't sure that it would be true in all cases.

The one that crossed my mind was the pl/sql array insert looking something like:

    for all i in 1..m_ct

        insert into tableX values(m_array1(i), m_array2(i) ...);

It operates as a genuine array insert - which means that /*+ append */ may be relevant - but it does have a values() clause - which means that /*+ append */ may be ignored.

Does your comment apply in this case as well as the single-row case ? I didn't have time to run up a test case to check.

-- 
Regards

Jonathan Lewis
http://jonathanlewis.wordpress.com

Author: Cost Based Oracle: Fundamentals
http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html

The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html