News for hackers. Stuff that matters.
Re:Getting 100x better search performance on Riak by gfodor on Sunday March 18, 2012 @06:59PM (#3721317)
The deficiency with AND queries due to term based partitioning reminds me of a hack I did a few years ago trying to build a basic IR system on simpledb:

http://codingthriller.blogspot.com/2008/04/simpledb-full-tex...

Of course, I was backed into this corner by the various constraints at the time, and this was just a fun hack. It sounds a bit precarious to consider using Riak for text search, at least compared to something like Solr, if it truly has issues with the basic bread and butter of search like sorting by recency and conjunctions. (of course I am only going on what is in this post, I have no experience with Riak)

Re:Getting 100x better search performance on Riak by gwf on Sunday March 18, 2012 @07:21PM (#3721383)
It really isn't all that bad. When you compare what we had to do compared to keeping two big data systems in sync (instead of Riak, which is a breeze), I think we got off easy.
Re:Getting 100x better search performance on Riak by nosequel on Sunday March 18, 2012 @08:38PM (#3721643)
Showyou designed Mecha to run Solr with Riak, there are slides here http://www.slideshare.net/jmuellerleile/scaling-with-riak-at... and I noticed someone else posted the video of the talk.