Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Efficiency of sqlldr

Re: Efficiency of sqlldr

From: Daniel Morgan <damorgan_at_x.washington.edu>
Date: Wed, 19 May 2004 22:27:42 -0700
Message-ID: <1085030866.302917@yasure>


Nan wrote:

> Hello -
>
> I have 20 database tables. The table spaces are configured on
> different disks.
> I have 10 Million rows of data for each table.
> (20 files with 10M rows each per table)
>
> Load Option #1:
> ---------------
> Use Oracle sqlldr (direct path) for all the 20 files with 10M rows of
> data each, into the respective table. One table is loaded after
> another.
>
> Load Option #2:
> ---------------
> Create batches of files, say 10,000 batches each with set of 20 files
> containng 1000 rows per file. Then I sqlldr the batches of files. One
> batch is processed after another. The 20 tables are populated in
> batches of 1000 rows each.
>
> Inference:
> ----------
> Does option #1 or option #2 make difference in terms of the efficieny
> of the queries that will run on the tables once data is loaded ?
>
> Which is more efficient ? I am looking at the performance of queries
> that run once the data is loaded through different options.. not
> necessarily the time it takes to load the data into the tables.
>
> Thanks
> Nan.

Given that we hae no informatoin on the time/effort/cost of "create batches of files" the only direction we can possibly point you is #1. Which would be my first choice even if the cost of #2 was zero.

-- 
Daniel Morgan
http://www.outreach.washington.edu/ext/certificates/oad/oad_crs.asp
http://www.outreach.washington.edu/ext/certificates/aoa/aoa_crs.asp
damorgan_at_x.washington.edu
(replace 'x' with a 'u' to reply)
Received on Thu May 20 2004 - 00:27:42 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US