Re: map/reduce question

[Roger Schildmeijer]

Cassandra 0.6 has basic Hadoop map/reduce support (see CASSANDRA-342). You 
could also look at the example in contrib/word_count.

--Roger Schildmeijer
On 4 mar 2010, at 18.00em, Matthew Chambers wrote:

> Hi everyone, I've been tasked with researching noSQL databases for use at 
> Imageworks.  So far I've checked out Couch, HBase, and Cassandra.  One 
> question I have, and I've looked around for this but couldn't find anything, 
> does Cassandra have integrated map/reduce functionality or do you access data 
> purely by unique key?
> 
> -Matt 

Re: map/reduce question

[Jonathan Ellis]

0.6 has Hadoop map/reduce support (see contrib/wordcount for an
example) but this is more for analytics (small numbers of large
queries) than live web app style load.

Cassandra also has range queries (get_slice_range) -- performing a
slice across sequential rows, where you don't necessarily know all the
keys in advance -- which is suitable for low latency queries.

On Thu, Mar 4, 2010 at 11:00 AM, Matthew Chambers
 wrote:
> Hi everyone, I've been tasked with researching noSQL databases for use at 
> Imageworks.  So far I've checked out Couch, HBase, and Cassandra.  One 
> question I have, and I've looked around for this but couldn't find anything, 
> does Cassandra have integrated map/reduce functionality or do you access data 
> purely by unique key?
>
> -Matt
>

map/reduce question

[Matthew Chambers]

Hi everyone, I've been tasked with researching noSQL databases for use at 
Imageworks.  So far I've checked out Couch, HBase, and Cassandra.  One question 
I have, and I've looked around for this but couldn't find anything, does 
Cassandra have integrated map/reduce functionality or do you access data purely 
by unique key?

-Matt