63

It’s been well over 10 years now since Real Time Statistics were first made available by a PTF to DB2 V7. Originally they were provided as an optional feature but since DB2 9 they have always been collected and the data is now fully integrated into the DB2 Catalog. The original aim of RTS was primarily to provide data to assist with the management of Imagecopies, Reorganizations and other Housekeeping tasks. While that remains a primary reason for their existence and something every site should be using, the last few releases of DB2 have added valuable data to RTS which can be used for a variety of other purposes including performance analysis and even database index design. Are you getting the most value from the RTS data which you are burning CPU cycles to collect? Come to this session and understand how RTS data is collected, how you can manage and control this process and most of all how to make the most of this valuable data source.

Content: Starting with an explanation of Big Data and the Apache Hadoop ecosystem with key components and initiatives, complementing z Analytics by using Big Data analytics and Natural Language Processing (NLP), integration of BigInsights with z Systems, customer use case scenarios, keeping DB2 for z/OS as control point for real-time z Analytics with SPSS.

Abstract: During the past decade, the Apache Hadoop open source community has developed an impressive set of products and tools that represent an important aspect of Big Data. In the beginning, we will explain the term Big Data and its relationship to the Hadoop ecosystem. We will furthermore illustrate how to complement existing 'traditional' DB2 for z/OS based analytical capabilities with Big Data analytics, e.g. by using text analytics and Natural Language Processing (NLP) as part of IBM InfoSphere BigInsights. This leverages key Hadoop components (MapReduce programming model, HDFS, HBase, Zookeeper, etc.) to analyze data from Enterprise-owned systems of engagement (e.g. call center transcripts, e-mail traffic, Facebook), and data from external social media sites (e.g. Twitter tweeds, Facebook sites, Blogs) and putting this in context with transaction insight from data on IBM z Systems. We will provide examples on how Hadoop systems - by using HBase and Hive with corresponding connectors to existing z Systems – and Big SQL on HDFS and Hive will enrich analytical insight. We will describe the IBM SPSS Modeler Server Scoring Adapter for DB2 on z/OS that embeds scoring algorithms directly into IBM DB2 for z/OS transactional data, which enables the real-time scoring of new transactional data as it is created.

The past couple of DB2 releases has been interesting in terms og providing both Development and Database Administrator a wealth of features to do “autonomic computing” and to preserve development time getting new/changed applications to market faster than ever before. You “JUST” need to identify how you can take advantage of these COOL new “things” and how to make your favorite DBA your best friend. After all – making DB2 successful requires the application development teams AND the DBA’s to get together and work together. We will discuss some of the latest features related to how applications can take advantage of DB2 but also how these application changes need to be understood by the Database Administrators.