IBM Streams 4.2

Running applications in stand-alone or distributed execution environments

You can run your applications in either stand-alone or distributed execution environments in IBM® Streams Studio.

About this task

When an application is compiled, an application bundle file is generated along with a stand-alone executable file with the name standalone in the output/bin directory. IBM Streams supports two types of execution environments:
  • You can run your application in stand-alone execution environment. In stand-alone executions, an application is an executable file that you can launch directly. It does not run on the IBM Streams runtime environment. In stand-alone executions, all operators are fused into a single partition and a single processing element (PE) is generated for this partition. When the stand-alone executable file is launched, this executable file loads the PE.
    Note: SPL projects with a remote context are run on the remote system.
  • You can run your application in distributed execution environments. In distributed execution environments, an application is run when you submit an application bundle file to the IBM Streams runtime environment for execution. Unlike the stand-alone execution, the operators in an application that is launched in distributed execution environment can be fused to more than one PE, and the PEs can be distributed onto multiple hosts. To run an application in a distributed execution environment on a IBM Streams instance, you need to provide the compiler-generated application bundle file to the streamtool commands.

Streams Studio provides launch support for both stand-alone and distributed execution environments, regardless of how you build your application.