Post a Comment

InfoSphere Streams 3.2 Announced

October 22, 2013

It seems like a “Back to the Future” moment. Here we are with the IBM InfoSphere Streams v3.2 announcement, the latest version of our product for handling stream computing and complex event processing. Yet 5 years ago this month, we had IBM System S v3.2. Looking back, we had three manuals for documentation – a language reference, writing business applications document, and the Installation and Administration guide. There’s certainly a strong familial resemblance, especially with the programming language and the clustered runtime.

So, what’s happened in the intervening 5 years between these two offerings each denoted Version 3.2? A tremendous amount! IBM InfoSphere Streams v3.2 is the ninth release, dating back to Streams v1.0 which shipped on May 15, 2009. That first release eliminated the need for customers to download source code from IBM and also open source modules from other sites. Previously, it was like an open source project, since customers had to compile all the code together and then install them all manually – a far cry from commercial expectations for a simple install.

The next 7 releases added improved security, a revamped and standardized programming language, visual drag-and-drop development, sophisticated performance and metric monitoring in the administrative console, additional hardware and operating system support, XML data type support, and a huge number of toolkits and accelerators to speed development.

Toolkits included adapters to major databases, messaging systems, IBM InfoSphere DataStage and Hadoop File Systems (HDFS). Toolkits also included a large variety of analytic functions, including advanced text analytics for natural language processing, SPSS and R support, geospatial analytics and the time series toolkit for signal processing and machine learning capabilities. The telco and social media accelerators added complete applications for telco mediation and analysis of social media streaming data like Twitter, Tumblr and blog sites.

Streams v3.2 continues to deliver the same tremendous performance, reliability and usability that attract customers and partners. Performance tests routinely show the ability to handle ten times the volume of records as other streaming and complex event processing systems. The runtime stability has impressed many users, with some having applications put into production without a single failure in over two years. Streams Studio simplifies development usability with Eclipse-based, context-sensitive help and code assist to speed application development.

Streams V3.2 integrates with existing infrastructures and includes new features and functions designed to instantly analyze and fuse massive volumes of data. It delivers new capabilities to improve usability, support enterprise readiness, and streamline enterprise integration.

Perhaps the most exciting new addition for many users will be Remote Streams Studio for Windows and Linux: Use Streams Studio natively on a Windows desktop to develop Streams applications. Streams Studio interacts seamlessly with your Linux-based Streams cluster instances to test and deploy applications. End users will also appreciate the simplifications made to create new visualizations of streaming data in charts, graphs and tables.

New adapters allow for data exchange with the Internet of Things via the MQ Telemetry Transport (MQTT) protocol, implemented by many devices and in IBM’s MessageSight appliance. Another adapter allows passing records to IBM Operational Decision Management (ODM) which evaluates the records based on its rules engine, and passes results back to Streams. Database support for IBM InfoSphere BigInsights BigSQL, and ANSI standard SQL running on our Hadoop-based offering, and also for HP Vertica.

For those few government and early adopters of System S, like University of Ontario Institute of Technology in Ontario, Canada, and Uppsala University and KTH University, both in Sweden, and Beacon Institute in the United States, they will continue to recognize the advanced streaming analytic engine and tools due to the ancestry of System S v3.2. But the return to today will continue to impress them with the market-leading capabilities of Streams to deliver real-time analytic processing applications.