Apache Spark for zOS NOW available on developerworks [ Link: ibm.biz/BdHbsZ ]

3
© 2010 IBM Corporation Apache Spark The IBM z/OS Platform for Apache Spark has just gone live on developerWorks ! 1

Transcript of Apache Spark for zOS NOW available on developerworks [ Link: ibm.biz/BdHbsZ ]

Page 1: Apache Spark for zOS NOW available on developerworks [ Link: ibm.biz/BdHbsZ ]

© 2010 IBM Corporation

Apache Spark

The IBM z/OS Platform for Apache Spark has just gone live on developerWorks!

1

Page 2: Apache Spark for zOS NOW available on developerworks [ Link: ibm.biz/BdHbsZ ]

© 2015 IBM Corporation2

What is Apache “Spark”?

Apache Spark is an open source cluster computing framework based for in memory transactions

It can run faster than Hadoop, due to it being in memory rather than in disk

It can interact and link to Hadoop clustered file system, Cassandra, Openstack Swift – for storage connectivity (so can act as the in memory processing platform for large scale big data analytics)

Spark had in excess of 465 contributors in 2014... Not only the most active project in the Apache Software Foundation,

but one of the most active open source big data projects

Page 3: Apache Spark for zOS NOW available on developerworks [ Link: ibm.biz/BdHbsZ ]

© 2015 IBM Corporation3

What are IBM offering...

IBM packages for Apache Spark an integrated, highly performant, and manageable Apache Spark ™

runtime, tuned for solving analytics problems on IBM platforms. • z/OS• Red Hat • Suse

Keep's Mainframe relevant as processing platform for in-memory Spark opportunities

• Allows us to have the conversation around in-memory opensource analytics

IBM Unique is: • the ability to develop Apache Spark runtimes for z/OS workloads either

on z/OS (or on Suse/Red Hat Linux) • z13 performance on Spark is significant vs. x86 (unlike Hadoop)

3