Re: sqlldr very very slow

freCho wrote:

> Hi all,
> I'm having problem using sqlldr.
> I have to load about 6.000.000 records into a large table
> (about 300.000.000 records). This mornig I started a script which
> connect to db and launch sqlldr....but I'm waiting for the FIRST
> "Save data point reached" !!
> The db is an Oracle9i Enterprise Edition Release 9.2.0.1.0,
> the server is a Sun Enterprise 420R whit 4 CPU and 4GB RAM
> Solaris 8.
> The control file used by sqlldr utility is this:
>
> options (direct=true, rows=100000)
> unrecoverable load data
> infile 'first.csv'
> append
> into table TABLENAME
> fields terminated by ','
> trailing
> (
> column names here
> )
>
> The db is out of service, so only background process and the sqlldr are
> connected to it:
>
> SQL> select username, sid,MACHINE,PROCESS from v$session;
>
> USERNAME SID MACHINE PROCESS
> ------------------------------ ---------- ------------------------------ ------------
> 1 dbadv 3813
> 2 dbadv 3815
> 3 dbadv 3817
> 4 dbadv 3819
> 5 dbadv 3821
> 6 dbadv 3825
> 7 dbadv 3827
> 8 dbadv 3829
> ENG 191 dbadv 22554
> SYS 259 dbadv 16507
>
>
> After 3 hours waiting I started a little test on another server running
> an 8i.
> I created the same table, and using the same input file and control file
> (changing only the connect string of course), all records are been loaded
> in about 6 mins !!!
>
> There are differences between the two schema used in the different oralce server, I have not created
> all tables related with the original one, so no time have been spent in
> the test db to check constraints, but I don't think that this can be the root cause of the
> slowness in the 9i environment.
>
> What can I check to see what is happening??
> If same other information is required please ask me.
> Any idea is very appreciated.
>
> ps- sorry for my poor english

kill -9

Then try again using one record.
Then 10
Then 100
to sort it out.

-- 
Daniel A. Morgan
University of Washington
damorgan_at_x.washington.edu
(replace 'x' with 'u' to respond)