I'm an advocate of using Jython for BigData eco-system integration projects.
The main reason is development velocity and productivity clearly established by Python.
Additional justifications...
-
Ease of access to all Java library packages. Thus, no vias such as thrift are required.
-
For long running programs, Python on the JVM is performant.
-
Jython can leverage existing CPython modules.
-
Script, test, script, test... no long compiles and builds required.
Drop me a line if you find any of this useful, or if you have any constructive comments or insights.
## Contents
* jython-shell-script/ * Jython shell scripts that make it easy to include various BigData eco-system jars. * Simplify access to HBase.
-
lib/
- Libraries to access HDFS, HBase, Elasticsearch, etc.
-
conf/
- Configuration file examples.
-
tests/
- Run tests for the lib/ modules.
-
examples/
- Working example Jython code.
## Features.
* Boon JSON processor
-
HBase
-
HDFS
-
Properties Configuration Object
-
Worker Threads Model
## Planned...
* more examples/...
-
lib/Tachyon.py
-
lib/0mq.py
Thanks,
-dsidlo