(c) Copyright 2013 WibiData, Inc.
This project contains the skeleton for a web application that uses the Kiji platform. It contains two modules in this directory:
- lib - Packages a JAR of gatherers, producers, importers, etc.
- app - The web application servlet.
You must have set HADOOP_HOME and HBASE_HOME environment variables to allow the web application to work with Kiji. In particular, $HADOOP_HOME/conf/{core,hdfs,mapred}-site.xml files need to specify the addresses of your Hadoop cluster, and $HBASE_HOME/conf/hbase-site.xml needs to specify the address of your ZooKeeper/HBase cluster.
The easiest way to run this application is to download the Kiji BentoBox
and run the bin/kiji-env.sh script in the BentoBox directory, and then
run bento start.
The application requires some amount of data processing to be done in batch before the webserver is started. This can be done on a one-time basis by running the build.sh script in the top-level directory.
$ ./build.sh
Once the app and lib modules have been successfully built, the Kiji schema has been created, and all of the batch processing has been completed, the webserver can be started by running the start.sh script in the top-level directory.
$ ./start.sh
Open up a web browser to http://localhost:11786/ to see your app in action!
The "app" module depends on the lib project. To build the entire project, type 'mvn install' at the top level (in this directory).
$ mvn install
If successful, you can launch the web application on your local machine using the maven jetty plugin. You must do this from the 'app' module.
$ cd app
$ mvn jetty:run