[Koha] redundant records and Error 404s

[Chris Cormack]

On 20 March 2010 10:01, Otis, Perry W Civ USAF AETC AUL/LTIA
<perry.otis at maxwell.af.mil> wrote:
> I experienced this same problem.  The problem is that Zebra still has
> those records indexed, even though they no longer exist.  Running
> rebuild_zebra.pl will not help.  I tried dropping the database.  I tried
> uninstalling and reinstalling Zebra.  I even tried uninstalling and
> reinstalling Koha.  Didn't work.  I looked around some more and finally
> found some files in /var/lib/koha/zebradb/, which is where I now suspect
> that Zebra stores its indexes.  I deleted the /var/lib/koha/ directory
> and reinstalled Koha again and the problem was solved.  There may be a
> better way to do it, but that is what worked for me.
>
There is :)

You can delete the zebra db (not the mysql one), and create a new one
using zebraidx commands, much better than deleting files and
reinstalling, here's how you do that.

As the user your zebra runs as (probably koha)
run
zebraidx -c /path/to/zebra-biblios.cfg drop biblios

For me that is
zebraidx -c /home/chris/koha-dev/etc/zebradb/zebra-biblios.cfg drop biblios

Then
zebraidx -c /path/to/zebra-biblios.cfg commit

Then you can run rebuild_zebra.pl -b -r -v

The -r should reset your index anyway, so you only need to drop the
zebra index if that didn't work.

Chris