IBM Streaming Analytics service

IBM® Streaming Analytics service is a cloud version of IBM Streams that is available from IBM Bluemix®. Streaming Analytics is a fully managed service that frees you from time-consuming installation, administration, and management tasks, giving you more time to develop streaming applications.

How do I sign up for Streaming Analytics service?

For information about the Streaming Analytics service plans, see the Bluemix catalog page.

What version of Streaming Analytics service am I using?

Improvements are pushed regularly to all Streaming Analytics services. You always use the latest version of the managed service, and there is no product version or level for you to keep track of.

What does IBM manage for me?

We handle installation, software upgrades, creating and managing domains, and hardware maintenance. The service includes 24 x 7 health monitoring.

What tasks am I responsible for?

You write the applications that will run in a Streaming Analytics service and Streams instance on-prem and ensure that they are functioning correctly and meeting performance requirements. You are also responsible for any application-specific monitoring.

Is high availability (HA) supported?

High availability is managed by IBM. Streaming Analytics is configured to support HA. Additional server resources are in place in the event of a failover.

How is security managed for the Streaming Analytics service?

Security is fully managed by IBM. Credentials are generated for each service and are provided to you. Security updates are managed and applied by IBM promptly after they become available.

Do I need to configure a Streaming Analytics service?

The service is created and fully managed by IBM. Each service consists of a dedicated set of application nodes.

How do I develop Streams applications?

You must develop Streams applications locally by using the free Streams IBM Streams Quick Start Edition or the IBM Streams Developer Edition. You can also use the on-premises IBM Streams installation if you have one. Applications that you develop and compile locally can then be seamlessly deployed as a bundle to a Streams service in the cloud.

But if you want to run your Python applications in the cloud, you don’t need to install IBM Streams on-premises. Simply use the STREAMING_ANALYTICS_SERVICE context to submit your Python applications to the Streaming Analytics service. You can develop the applications in a standard Python development environment or in a Jupyter interactive notebook, but you must use Python 3.5.

For guidance on developing applications, see the Bluemix Streaming Analytics Development Guide.

Can I sign in to a Streaming Analytics service host directly?

No. A direct login to the server with Telnet or a Secure Shell (ssh) is not supported. You cannot install additional software or run non-Streams software on a Streams host.

Can I access the file system on the Streaming Analytics service?

Only your Streams applications can directly access the file system on a Streams host.

Cloud-ready alternatives exist for solutions that need a mechanism for Streams to interact with other solution components. You can use other source and sink protocols instead of files. For example, cloud-based data stores.

How can the Streaming Analytics service applications access my organization's enterprise data?

You can use the Bluemix Secure Gateway Service to securely connect streams applications to your enterprise.

Are all the features for IBM Streams for on premises supported by the Streaming Analytics service in the cloud?

Some of the features that are not supported for Streaming Analytics service include the following:

• Administrative tasks for an instance that require domain authority. For example, adding custom host tags or creating a job group.
• Consistent region checkpoints.
• Some of the toolkit operators are not supported. For information, see Compatible operators.
• The Streams JMX API.

Where can I learn more about the Streaming Analytics service?

The IBM developerWorks® article, Roadmap for Streaming Analytics Service on Bluemix, has guidance for learning about the service, plus links to informative blogs, demos, and articles.