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

Home -> Community -> Mailing Lists -> Oracle-L -> RE: quickest method

RE: quickest method

From: Mark Work <mark_at_cool-tools.co.uk>
Date: Thu, 15 May 2003 10:31:46 -0800
Message-ID: <F001.00599F56.20030515103146@fatcity.com>


RE: quickest methodThis is in fact how DataBee, our Oracle subsetting tool, works. It identifies the ROWIDs of the rows required for the subset from the source database, then connects to the target database, and sucks the rows down, inserting via a select .. @dblink (server to server, so often on a highspeed backbone anyway)..

It performs very well. I'm sure Dale Edgar (DataBee's primary developer, and another list member) has done some tests on this at some point - did you Dale (if your watching)?

Mark
Cool-Tools UK Limited
  -----Original Message-----
  From: root_at_fatcity.com [mailto:root_at_fatcity.com]On Behalf Of Jankovic, Djordje
  Sent: 15 May 2003 18:22
  To: Multiple recipients of list ORACLE-L   Subject: RE: quickest method

  If the source and target are on separate servers sql*loader approach requires spooling data to a file, ftp-ing the file over, running sql*loader. Depending of the size of the data, direct load through database links (insert /*append*/ select ... from xyz_at_link) can be faster than that.

  Export/Import can be done through named pipes, even when source and target are on different servers. Anybody tried sql*loader through named pipes: I guess it is doable?

  Djordje
    -----Original Message-----
    From: Adams, Matthew (GECP, MABG, 088130) [mailto:MATT.ADAMS_at_appl.ge.com]

    Sent: Thursday, May 15, 2003 8:35 AM     To: Multiple recipients of list ORACLE-L     Subject: RE: quickest method

    the sqlplus copy command is not inheirantly slow.     I think it has a MAJOR dependance on the speed of     the disk farm and the speed of the network.

    We used it yesterday to move a 52 million row     table (about 7 gig) in about 52 minutes.     That's not bad. SQL*Loader may have done it faster,     but we were satisifed with the speed of the 'copy'     command.



    Matt Adams - GE Appliances - matt.adams_at_appl.ge.com     When someone says "I want a programming language in which I     only need say what I want done", give him a lollipop.

    -----Original Message-----
    From: Darrell Landrum [mailto:dlandrum_at_zalecorp.com]     Sent: Wednesday, May 14, 2003 8:00 PM     To: Multiple recipients of list ORACLE-L     Subject: Re: quickest method

    A minor addendum:
    Based on this info, I'd say that using the sqlplus copy command is     probably the slowest. In one scenario using sqlplus copy to copy some     tables, about 5 hours into what turned out to be a 12 hour process, I     started exporting the same tables, copied across to the target server,     and imported in less than 1/3 the time.     I don't have a lot of experience with SQL Loader, but in a few     optimized cases (using direct load), SQL Loader screamed.

>>> Jared.Still_at_radisys.com 05/14/03 06:01PM >>>
    Carol,

    Hands down, SQL Loader is the fastest.

    Export/Import is rather slow.

    SQL and PL/SQL commands can be on either side of exp/imp, depending     on what you are doing and how well the code is written.

    e.g. SQL statements are fairly fast, PL/SQL for loops are not. Pl/SQL

    bulk
    processing is fast.

    Unless you need the programatic abilities of PL/SQL, use SQL Loader.

    Exp/Imp can still be useful, even with SQL Loader. Use exp/imp to     build
    your tables, then the indexes and constraints after the data is     loader.

    No pat answer as to how to load data, depends on your requirements.

    There's probably no point in messing with SQL Loader if the data sets     are small, and you can easily export from another database and then     import.

    If the data is in CSV or flat files though, and/or is very large, SQL     Loader
    is very fast.

    HTH     Jared

    "Carol Legros" <carol_legros_at_hotmail.com>     Sent by: root_at_fatcity.com

     05/14/2003 02:57 PM
     Please respond to ORACLE-L


            To:     Multiple recipients of list ORACLE-L
    <ORACLE-L_at_fatcity.com>
            cc:
            Subject:        quickest method



    I'm curious to know whether anyone out there has seen a comparison     discussing the pros and cons and/or results of any simulation tests     that
    compare the speed with which data can be loaded into a target database

    from
    a source (database or flat file) using the following 3 methods :

    (i)   Export (from source), Import (to target)
    (ii)  SQL*Loader (to target)
    (iii)  SQL or PL/SQL commands (insert to target)
          using a Database Link between source &
          target

    I'm working on a data loading strategy and since there are "many ways     to
    skin a cat", I'm considering these as options. Of course, there are     other

    criteria that impact the method chosen, but assuming all things are     equal
    (ie network bandwidth is good, access to both source and target are not     an

    issue etc.), which of these methods would be quickest ?

    Thanks,
    Carol



    Add photos to your e-mail with MSN 8. Get 2 months FREE*.     http://join.msn.com/?page=features/featuredemail

    --
    Please see the official ORACLE-L FAQ: http://www.orafaq.net     --
    Author: Carol Legros
      INET: carol_legros_at_hotmail.com

    Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
    San Diego, California        -- Mailing list and web hosting services
    ---------------------------------------------------------------------
    To REMOVE yourself from this mailing list, send an E-Mail message     to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in     the message BODY, include a line containing: UNSUB ORACLE-L     (or the name of mailing list you want to be removed from). You may     also send the HELP command for other information (like subscribing).

    --
    Please see the official ORACLE-L FAQ: http://www.orafaq.net     --
    Author:
      INET: Jared.Still_at_radisys.com

    Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
    San Diego, California        -- Mailing list and web hosting services
    ---------------------------------------------------------------------
    To REMOVE yourself from this mailing list, send an E-Mail message     to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in     the message BODY, include a line containing: UNSUB ORACLE-L     (or the name of mailing list you want to be removed from). You may     also send the HELP command for other information (like subscribing).

    --
    Please see the official ORACLE-L FAQ: http://www.orafaq.net     --
    Author: Darrell Landrum
      INET: dlandrum_at_zalecorp.com

    Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
    San Diego, California        -- Mailing list and web hosting services
    ---------------------------------------------------------------------
    To REMOVE yourself from this mailing list, send an E-Mail message     to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in     the message BODY, include a line containing: UNSUB ORACLE-L     (or the name of mailing list you want to be removed from). You may     also send the HELP command for other information (like subscribing).
-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.net
-- 
Author: Mark Work
  INET: mark_at_cool-tools.co.uk

Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
San Diego, California        -- Mailing list and web hosting services
---------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).
Received on Thu May 15 2003 - 13:31:46 CDT

Original text of this message

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