On 21/10/11 17:42, Fred Zimmerman wrote:
Thanks for sharing this.
Some notes:
* Useless cats
You seem to prefer
cat enwiki-<date>.xml | mwimport > enwiki.sql
to
mwimport < enwiki-<date>.xml > enwiki.sql
While both commands do the same, and the overhead of cat can usually be
ignored, given that it's a gigantic file, I would still prefer to remove
that extra process.
* Temporary enwiki.sql file
You only seem to use enwiki.sql once, so I think you could have done
mwimport < enwiki.sql | mysql -f ...
and save another 33 Gb (assuming the mysql command won't fail :)
* Redirecting instead of piping
cat enwiki.sql > mysql -f ...
should have been
cat enwiki.sql | mysql -f ...
* SQL import instead of rebuildall.php.
The step 8 could have been skipped if you had imported instead the
*links dumps (although they are not completely in-sync with the xml).
* Blog entry to your inbox
Not related to the provided entry, but your entry
http://business.zimzaz.com/wordpress/2011/10/wikitech-l-5-zimzaz-wfzgmail-c…
contains just a link to your gmail inbox:
https://mail.google.com/mail/?shva=1&zx=wear6iy60ej8#label/Wikitech-l
(so it's useless for anyone else)