Received: by 10.50.12.169 with SMTP id z9mr205890igb.5.1332258706248;
        Tue, 20 Mar 2012 08:51:46 -0700 (PDT)
X-BeenThere: django-haystack@googlegroups.com
Received: by 10.182.46.136 with SMTP id v8ls336532obm.4.gmail; Tue, 20 Mar
 2012 08:51:44 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.68.233.131 with SMTP id tw3mr213036pbc.13.1332258704477; Tue,
 20 Mar 2012 08:51:44 -0700 (PDT)
Authentication-Results: ls.google.com; spf=pass (google.com: domain of
 wakeb0a...@gmail.com designates internal as permitted sender)
 smtp.mail=wakeb0a...@gmail.com; dkim=pass
 header...@gmail.com
Received: by pz2g2000pbc.googlegroups.com with HTTP; Tue, 20 Mar 2012 08:51:44
 -0700 (PDT)
Date: Tue, 20 Mar 2012 08:51:44 -0700 (PDT)
User-Agent: G2/1.0
X-HTTP-UserAgent: Mozilla/5.0 (X11; Linux i686) AppleWebKit/535.11 (KHTML,
 like Gecko) Chrome/17.0.963.56 Safari/535.11,gzip(gfe)
Message-ID: <0c08da89-dc97-4623-88b1-d38172e297bb@pz2g2000pbc.googlegroups.com>
Subject: Updating/rebulding index effects other solr cores
From: Dave <wakeb0a...@gmail.com>
To: django-haystack <django-haystack@googlegroups.com>
Content-Type: text/plain; charset=ISO-8859-1

I have two sites that are running on a multi-core solr setup.  They
both use the same code base (but have separate databases, similar
products).  When i rebuild or update the index on one site, it effects
the other site.  Any ideas whats going on?

When i update or rebuild, I get the correct model object count from
each site (it says something like 'indexing X products now, etc).

I believe whats going on is:
If i update/rebuild on site A, then the matching products between the
two sites are updated/rebuilt on site B...  so if site A has 100
products, 20 of them match on site B, then site B will only have those
20 products, and same happens reverse.

I can't tell exactly if that is what is going on due to the large
number of products (and a large number of similar products between the
two sites), but I'm pretty sure that is whats happening.

I'm mainly seeing this because site A has a very nice facet structure,
where site B does not.   If i update site B, my facets go to hell on
site A (because there is a small number of facets shared between the
two sites).  If I update site A, then the facets that site B shares
with site A are updated.

Ideas?
Thanks!
Dave