Just in case you haven't heard the news, IBM has released the "IBM Packages for Apache Spark," available for download at no charge here:
https://www.ibm.com/developerworks/java/jdk/spark/ Borrowing a little from Wikipedia, Apache Spark(TM) is an open source cluster computing framework well suited to solving many types of big data analytics problems, and it also supports certain types of machine learning algorithms. Apache Spark is available for several platforms including especially IBM z Systems. Both z/OS and Linux on z versions are available. If you'd like some more information on typical use cases for Apache Spark on IBM z Systems, this YouTube video should be helpful: https://www.youtube.com/watch?v=sDmWcuO5Rk8 Some more introductory information on Apache Spark is available here: http://www.ibmsystemsmag.com/mainframe/Business-Strategy/BI-and-Analytics/dynamic-Spark/?page=1 For z/OS you should have z/OS 2.1 or higher and the latest 64-bit IBM Java SDK (Version 8). It's possible Apache Spark will run on z/OS 1.13, but IBM has not tested that particular combination. If you don't have the latest Java SDK you can get it here: http://www.ibm.com/developerworks/java/jdk/ You also may need to download an updated version of bash. The installation documentation explains how to do that. Check back periodically for updated documentation that will detail more installation and configuration scenarios such as deployment to a Parallel Sysplex. Enjoy! -------------------------------------------------------------------------------------------------------- Timothy Sipples IT Architect Executive, Industry Solutions, IBM z Systems, AP/GCG/MEA E-Mail: [email protected] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
