Apache Hadoop integration to existing Architectures

This section suggests different ideas on the way we think is the best to integrate Hadoop to existing architectures.

How to integrate existing data with Hadoop?

be transparent in the data importation to DFS
user periodic batch import
asynchronic saving of data to DFS
take data from xml, cvs, data bases
Pentaho: desktop app ETL using Hive chroned using OS jobs. Full version includes importing from SAP repositories
Use Hadoop Java API to asynchronously updload files to the File System
Use Hadoop Distributed File System’s commands to import data asynchronously

Integration with DB

http://www.cloudera.com/blog/2011/06/biodiversity-indexing-migration-from-mysql-to-hadoop/

Log files processing

https://ccp.cloudera.com/display/CDHDOC/Mountable+HDFS

http://wiki.apache.org/hadoop/MountableHDFS

Visualization

Build an open source tool
Flashless tools:http://raphaeljs.com/ It doesn’t include the exportation to pdf!
Incorporate graphics to Rocketui de Globant. Rochetui includes widgets on Yahoo-ui, Prototype and JQuery
Google Visualization: there is a tool which obtains a static image as a printscreen by calling a URL
Lab de Adobe project to export Flash to HTML5
Fusion charts http://www.fusioncharts.com/
Info Build Toolkig
GWT: data models with canvas. GWT exporter can be used to be referenced with a JS application

Visualization Tools Comparison

http://sixrevisions.com/javascript/20-fresh-javascript-data-visualization-libraries/

How to generate Dashaboard once the output is ready?