Overview

Research Interests

  • Web Services
  • Grid Computing
  • WS Message Brokering
  • WS-Reliable Messaging
  • Workflow Composing

Projects

  • Apache Axis2 -I was a member of the pioneering Apache Axis2 team. The Axis2 project is the successor of the Axis SOAP project, which can be categorised as a third generation SOAP engine which was designed to be highly extensible so it can provide a framework to develop the WS-* specification(WSRM, WS-Security, etc) . It gave birth to a new StAX based infoset model, AXIOM, which leverages the concept of differed building to greately improve the memory foot print while giving a JDOM like API for convienient programming.
  • Apache EWS (Enterprise Web Services) - Apache EWS is JSR 109 implementation which in substance defines how the SOAP stack should be integrated to J2EE application servers. EWS was the JSR 109 implementation for Apache and it is currently used in Apache Geronimo and JOnAS Application Servers. This was part of my final year project for the B.Sc.
  • WS-Messenger - WS-Messenger is a highly scalable WS-Brokered Notification and WS-Eventing implementation developped at the Extreme computing Laboratory. Further it is capable of providing mediation between the two Web Service publish/subscribe inplementations that it support. It is scalable such that it can be run in multiple maching for improved throughput or load balancing. I have contrebuted to improve the scalability of the WS-Messenger and I am currently working on providing a WSRM implementation to WS-Messenger to make it support end to end reliable delivary of Notifications/Events.
  • LEAD -  Linked Environments for Atmospheric Discovery is a project that is funded by NSF which is geared to build a distrebuted infrastructure that can acuurately analyse, simulate and predict atmospheric changes and thus providing means for better meteological forecasting. This is project involved in IU,NCSA,OU, UniData and i am collaborating with the team at IU Extreme Computing Laboratary to port LEAD to Goobus Toolkit 4.
  • X-Q - X-Q is a java based distrebuted persistant Queue implementation done as a sub project under WS-Messenger and currently lies in the WS-Messenger CVS. It can be run in a cluster an infact the high scalable nature of the WS-Messenger greatly appreciated because it uses the X-Q internally. It has been tested for over one million items(size-Notification message with headders) of queue length and could operate at over a 500 items per second. The Queuing rate dose not vary as the Queue size grows.

Publications


Chathura Herath on March 06 2006