Solr Indexing and required fields.

Ryan Zezeski rzezeski at basho.com
Tue May 22 14:35:00 EDT 2012


On Tue, May 22, 2012 at 11:44 AM, Kevin Hill <virtual.khill at gmail.com>wrote:
>
>
> Interesting.  I built something like that as a prototype - using Lucene
> libraries in Java to index data in Riak and store the Riak keys as part of
> the indexed documents.
>
> The prototype involved storing product data in Riak and retrieving it
> directly via key (based on unique SKU identifier) or performing complex
> queries on product data in Lucene and loading data based on the Riak keys
> in the search results.
>
> Is that the same type of functionality you're talking about?
>
> - khill
>
>
Kevin,

That's not quite what I was talking about but that is certainly something
that could be done.  John Muellerleile (@jrecursive) has actually done work
to merge Riak/Solr but IIRC the project, named Mecha, is not in the wild
yet.

What I'm really talking about is getting the Solr emulation out of Riak
Search itself and into a self-contained application that can optionally be
deployed with Riak.  I'm of the opinion that full-blown Solr compatibility
is not something that belongs in the core of Riak.  It's too much to try to
maintain and keep up with and is not the main focus of Riak.  I'd rather
see it as an optional thing on the side and allow Riak to focus on it's own
search/index story rather than play feature checklist with Solr.

-Z
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.basho.com/pipermail/riak-users_lists.basho.com/attachments/20120522/8f2f57c2/attachment.html>


More information about the riak-users mailing list