I like the feedback, I had forgotten all about LOAD DATA INFILE.  I'd like
to focus more that this is for any large file processing, not only just
database.
Keep the feedback coming.  I just finished putting 79m apache log records
into the system.  I'll duplicate this using the load data setup and see how
fast it is.

--Will



On Wed, Sep 2, 2009 at 3:38 PM, Nicholas Leippe <[email protected]> wrote:

> I totally agree, and have used this technique many times for data imports.
>
> Data scrubbing can be messy, but once it's in a table, SQL is so powerful
> it
> can make an initially daunting-in-appearance task very manageable.
>
> As a worst case scenario, you may need to use sed/awk/perl/(or gasp, maybe
> php) to filter the file if it's not properly delimited. But it's often
> _much_
> faster to filter the file on disk first and then use a LOAD DATA INFILE
> than
> to even batch it in via INSERT queries.
>
> I haven't tried this, but you might be able to save yourself the extra copy
> if
> you need to filter it this way by writing the filter output to a named pipe
> and LOAD DATA INFILE on the pipe...
>
>
>
> /*
> PLUG: http://plug.org, #utah on irc.freenode.net
> Unsubscribe: http://plug.org/mailman/options/plug
> Don't fear the penguin.
> */
>



-- 
Take care,
William Attwood
Idea Extraordinaire
[email protected]

Jonathan Swift<http://www.brainyquote.com/quotes/authors/j/jonathan_swift.html>
- "May you live every day of your life."

/*
PLUG: http://plug.org, #utah on irc.freenode.net
Unsubscribe: http://plug.org/mailman/options/plug
Don't fear the penguin.
*/

Reply via email to