Using pipes

Pipes can be used whenever data is transferred using sequential data sets. Pipes are identified by a pipe name that adheres to data set name rules. A pipe name must be unique (only one pipe with a given name can be active at any time). When a pipe replaces a data set, the pipe name is the name of the data set that it replaces.

Pipe participants must run in parallel. They can run in the same system or in different systems in a Parallel Sysplex environment.

The writer of the pipe can optionally create a file that contains the data that has been written to the pipe. This file creation is performed by the Job Optimizer Pipes Subsystem and is transparent to the application and, optionally, to the JCL.

Pipes can be implemented by the following methods:

  • Pipe Rules
  • JCL or dynamic allocation, by using the SUBSYS parameter to reference the Job Optimizer Pipes Subsystem

A job can refer to any number of pipes in the same step by using any combination of these methods.

This section provides the following topics: 

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*