Open Stack

Hi All,

I would like to share with you about my recent efforts on the integration 
between Swift and Apache Spark. 
Spark  claims to have x100 faster map reduce analytics than conventional 
Apache Hadoop. (http://spark.apache.org/ for more information about Spark)
Spark can read data from various sources, HDFS, S3, local file system, 
various streaming sources.  Spark then used to perform analytics on this 
data. 

I started to work on the integration between Spark and Swift, allowing 
Spark to integrate with Swift and perform data analytics on the objects 
stored in Swift. In my local tests this works very well. There are no 
modification needed for Swift.  I submitted patches to the Spark community 
with information how to integrate it with Swift. This work is still in 
progress.

https://github.com/apache/spark/pull/1010


All the best,
Gil Vernik.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20140614/2ae3a749/attachment.html>