Re: massive delete

> I am not quite clear about your reply :-(

Excuse me, English isn't easy for me O:-)

>1. Is the solution no 1 helpful to you or not ?? The only idea behind
using

Yes, it is. (thanks again :-)
Finally I've done something like this:

select count(*)

    into :n
    from table_1
   where field1= xxxx and field2= yyyy;

for( i=0; i<n/10000 + 1; i++ )
{

    delete table_1 where field1= xxxx and field2= yyyy AND ROWNUM<10001;     commit;
}

>2. Truncate is a DDL command that frees all extents from the table and
>drops the high-watermark too, thus freeing up space which would probably
>lead to fragmentation on your disk.

Ok. what about dropping the table instead of truncate it to avoid fragmentation?
(suppose a table without constraints)

thanks

    Jorge Received on Fri Sep 24 1999 - 05:32:57 CDT