Standalone example on jberet (jsr352)

[nageswara.raov]

Is there anyway to use jberet as standalone module to execute Batch Jobs?

Here is the usecase

-----------------------------

- I have 5 stages of data processing

- A batch job is to be executed at each state

     - batch job consists of reader, processor, writer

- Each stage is has to wait for the completion of its former

- Data flow is required between stages

All the steps in use-cases are perfectly in-line with jBatch specification. Fortunately or Un-Fortunately it has to be executed as a standalone data processor.

On search getting only WildFly samples. Not standalone all the time.

Any insights on why/why not would be helpful.

[csscouter]

You may find this example helpful, if you've not already seen it: http://www.mastertheboss.com/javaee/batch-api/running-batch-jobs-in-j2se-applications.

With some effort, I got it to run. I have not been able to translate its execution under Maven (as in "mvn exec:java") to a standalone .jar file that I can execute on a command line. If that is the nature of your request here, then I'm in exactly the same spot you are (and thereby second your request for assistance here...).

[cfang]

Project JBeret contains some test apps in Java SE environment:

jsr352/test-apps at master · jberet/jsr352 · GitHub

There are several such test apps that focus on chunk processing: chunkPartition, chunkStop, chunkClose, etc.  And all infinispanRepository* test apps also use chunk processing.

The README at JBeret project page (jberet/jsr352 · GitHub) has some quick info how to run it in Java SE.  You may also want to check out http://jberet.gitbooks.io/jberet-user-guide/