2006 - 2009- Analyzed the source code and the theory behind Apache Hadoop (a Java distributed computing framework inspired by Google?s Map-Reduce and Google File System) and Apache HBase (a column-oriented distributed database similar to Google?s BigTable), and was responsible of explaining them to the team members.
Attended to ApacheCon US 2008 at New Orleans, Louisiana, USA.
- Fixed bugs and added 5 important new features and improvements to ETRI internal projects based on Hadoop and HBase, testing on a cluster of 40 machines, using the TDD (Test-Driven Development) methodology, a version control system (SVN), a Continuous Integration server (Hudson) and a bug-tracking tool (Mantis).
- Designed, implemented and tested a query language similar to SQL (including Select, Insert, Update and Delete statements) adapted to column-based databases like HBase using JavaCC (an open source parser generator).
- Wrote a white paper and presented it at KDBC 2008 (Korean DataBase Conference in Korea Polytechnic University, Ansan, South Korea), about the Intermediate Merger, an improvement to the Map-Reduce framework.
- Designed and developed a server application (in Tomcat) that exposes the EPCIS (Electronic Product Code Information Services) API as a web service (using Apache Axis), allowing client applications (including a web-based prototype client) to store and query EPC-related data into an XQuery database.
2006 - 2006Prototyped client and server applications (using Java J2EE, OSGi, JSF, JSP, Oracle 9i database, WAP, IBM WebSphere Application Server) that allow a user to order a hotel room using either a phone, a touch tablet in a car or a website, and open the door of the car and the hotel room using an NFC-enabled phone (Near Field Communication).