July 29, 2015 | Written by: Prachi Snehal
Categorized: What's New
Share this post:
IBM Streaming Analytics is the latest addition to the IBM Bluemix Platform. Now in open beta, Streaming Analytics enables you to instantaneously analyze data, scaling to thousands of sources, on the cloud. Streaming Analytics has connectors to leading analytics solutions such as Microsoft Excel, which enable you to visualize data easily and efficiently at the speed of business.
Find the Streaming Analytics service in the IBM Bluemix catalog. Also, read about how to use the service in this article.
Streaming Analytics is a fully managed service, which allows you to build streaming applications with ease, without having to worry about managing and maintaining infrastructure. It is completely scalable, dynamically building out the infrastructure as needed, giving developers time to focus on building important business logic and analytics. The service supports real-time analytics with extremely low latency and high performance. Whether your application supports a single device and data source or connects and monitors hundreds of thousands of devices, Streaming Analytics performs seamlessly and reliably.
Here is the typical flow of a streaming application:
Below are some examples of real-world applications of Streaming Analytics:
- IoT (Internet of Things) – From connected cars to thermostat regulation, Streaming Analytics can be used to monitor devices. It supports high throughput from hundreds of sources simultaneously, with sub-second latency.
- Consumer marketing – With Streaming Analytics, you can monitor customer location and push specific advertisements to devices in the vicinity of your establishment (also see the Geospatial Analytics service on Bluemix), gain insights into social media postings and tweets, and realize smarter campaign management. (Streaming Analytics has the capability to process delayed or unordered data.)
- Event Analytics – Wimbledon and the Tour de France leveraged IBM Streams to provide real-time statistics on important event metrics (for example, service speed, cycling speed, up-to-the-millisecond average speeds, etc.) Another possible application of Streaming Analytics is the analysis of online gaming through the ingestion and filtering of data sent by hundreds of game engine servers.
- Sensor data and equipment monitoring – Streaming Analytics can also be used for real-time demand forecasting of inventory and for early detection of machine servicing needs based on the initial onset of an anomaly.
- Enhanced security intelligence – Predict, prevent, and act on security threats using real-time fraud detection. Increase situational awareness and accelerate the resolution to malicious threats. Resolve complex security threats with real-time network monitoring. Learn how to predict, prevent, and act on security threats at the right time, and how to integrate predictive models with data streams.
Developing these types of applications is easy and can be done in multiple ways. Streaming Analytics supports a Java Application API, which means any Java developer can crank up an application with extreme ease. See the streamsx.topology Getting Started Guide on GitHub. You can also develop your applications both interactively (through the Streams Studio) and programmatically (through the Streaming Analytics REST API). If you need an environment to develop your IBM Streams application, you can start with the IBM Streams Quick Start Edition, free of charge.
Streaming Analytics is a stand-alone product supporting automated high availability, guaranteed tuple processing, and multi-tenancy. We have over 50 toolkits available, helping you connect your streams application with a wide variety of data sources. In addition, we provide a set of advanced analytics toolkits out of the box (for example forecasting and machine learning in time series, text analytics, geospatial, etc.), so that you do not have to build these capabilities from scratch. Finally, one of the most important advantages is the capability to integrate with other Bluemix applications seamlessly. This enables you to build end-to-solutions using numerous technologies all from the IBM Bluemix platform. For example, Streaming Analytics can connect to the Spark service (Apache Spark Beta). See the Bluemix catalog page to explore other services.
With the Streaming Analytics Beta offering, you get:
- Fault tolerance
- High Availability
- Disaster Recovery
- Shared management infrastructure with dedicated application resources
- Single domain to support hundreds of small instances along with the ability to add domains when needed
- Resource management/metering
- Automatic allocation of application resources as needed
- Automatic starting of your Streams instance
- Automatic stopping, after three hours, of any idle instances. However, you can restart your instance using the Service Admin Dashboard, the Streams Console, or the Service REST API
- Free support
- For any issues, you can submit questions on the forum. Our support team will monitor and respond to your issues/questions promptly
Sample applications, developerWorks articles, videos, and tutorials are provided to help you make the most of the Streaming Analytics service’s capabilities.
The Streaming Analytics service is in open Beta stage currently, which means you can use all these great features for free. Give it a try and let us know what you think!!
Please reach out to me on Twitter (@PSnehalP) if you have any questions or post a comment below.