Implementation of In-memory Data Grid (IMDG) in PST.AG


PST.AG handles a massive amount of data, for which it requires several state-of-the-art tools and methods. This research project, under Mr. Joydeep Bhattacharjee of PST.AG is to enhance the data management system. The main objective of the project is to manage, distribute and provide the back-up for the prepared data using a balanced structure. To carry on with the process, we are implementing Data Grid with multiple data nodes with master and client node to be released in multiple servers.  We are using Massively Parallel Processing (MPP) framework so that the future data releases will be done across the large cluster of servers.


Once the project is complete, we expect large improvements in productivity and performance with Deadlock-Free Transactions and Distributed In-Memory Transactions. The client node is designed to work in Round Robin (RR) fashion and with help the Self-Healing Cluster. This will allow us to configure multiple backup copies to ensure data resiliency and security in case of node failures.