A Gentle Introduction to JBatch

As Java EE 7 rolls forward, my good friend and colleague Arun Gupta has been trying to blog about some of the most exciting new features. His blog entry Undertanding JSR 352 Concepts is a great starting point into JBatch (aka "JSR 352" aka "Batch Applications for the Java Platform"). As you may be aware JBatch is an IBM led JSR that's slated to be included into the Java EE 7 platform, but not the lighter weight Web Profile.

Arun explains some key JBatch concepts like Job, Step, JobOperator, JobRepository and the Reader-Processor-Writer pattern, provides some pretty good code examples and points to further resources like the JavaOne JBatch session and the official spec artifacts.


how is this better than Quartz?

Posted by guest on February 07, 2013 at 07:15 AM PST #

There are certainly some similarities (for example, I've personally created pretty decent batch processing solutions on enterprise schedulers like Flux and Quatrz). The focus is really entirely different though.

While JBatch focuses on features important to batch processing, enterprise schedulers like Quartz are focused on UNIX cron style scheduling concerns. If you are familiar with it, JBatch more closely mirrors Spring Batch (SpringSource was a major committer on the JSR) or IBM's WebSphere Compute Grid (WCG) and z/OS Batch.

As you dig into more details beyond Arun's very basic intro, you'll see notable differences in things like chunks, parallel processing, flow, split, join, transactionality, retries, sequencing, partitioning and so on that are specific to robust batch processing as opposed to generic scheduling.

Hope this helps.

Posted by Reza Rahman on February 07, 2013 at 08:20 AM PST #

Notify me email

Posted by guest on February 08, 2013 at 02:36 AM PST #

Post a Comment:
Comments are closed for this entry.