Searching and sorting with mapreduce or solr?
odie.miranda at gmail.com
Mon May 2 20:30:10 EDT 2011
Thanks for you reply.
What would you suggest on using, the solr interface or the map-reduce?
What would be best to use when you are processing lets say 1 million
On Tue, May 3, 2011 at 7:13 AM, <siculars at gmail.com> wrote:
> Riak m/r is a scan operation. Solr is an indexed operation. By definition
> an indexed operation will be faster than a scan operation. Just something to
> keep in mind.
> You also mention "search by field" but Riak is opaque to values so I will
> just assume your values are json (not to say that you can't operate on non
> json data).
> Basically you are going to run a map function that has your filter in it.
> And then pass those results to a reduce function that implements a sort.
> Your actual implementation will take advantage of features made available
> which has very good support for building m/r queries.
> I have two articles on my blog which may help you out.
> Sent from my Verizon Wireless BlackBerry
> -----Original Message-----
> From: Odelon Miranda <odie.miranda at gmail.com>
> Sender: riak-users-bounces at lists.basho.com
> Date: Tue, 3 May 2011 06:41:40
> To: <riak-users at lists.basho.com>
> Subject: Searching and sorting with mapreduce or solr?
> riak-users mailing list
> riak-users at lists.basho.com
We stretch out both of our arms...
Pass through the clouds, straight to the sky...
Even though we touched the Moon and the Mars...
We still cannot touch the TRUTH.
--- from the manga BLEACH
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the riak-users