Spotlight on Prof. S.N. Maheshwari, May 2017

Mr Rakesh Khurana, Chief Architect for Passenger Reservation System at Center for Railway Information Systems (CRIS), talks about collaborating with Prof S N Maheshwari in creating a single nationwide train reservation system

25th May 2017


Bio
Rakesh Khurana is an accomplished Technology Leader with a proven track record of successfully taking on challenging responsibilities of implementing large scale and complex IT Systems for both Government and Corporate Sectors. He successfully led and managed the development, implementation and operations support of the Indian Railways Passenger Reservations System. He is currently in the role of a Global CIO for a very large, global Tier 1 Auto Ancillary manufacturer which supplies parts to practically all leading Automotive OEMs of the world. Mr Khurana received his BTech from IIT D in 1983 and an MTech in CS&E from IIT D in 1991.


1. Can you tell us in brief what the scope of the railway reservation project was, what your role in it was.

The Passenger Reservation System (PRS) of Indian Railways was developed and deployed in 1985 by the Computer Maintenance Corporation (CMC). The system was implemented at five different Server locations across India - New Delhi, Mumbai, Secunderabad, Calcutta and Chennai, each managed and operated by a separate geographical division of the Indian Railways. These were five independent stand-alone systems catering to reservations for trains originating from within the geographical region associated with the server where the system was installed.

In the early 1990s the Center for Railway Information Systems (CRIS) was charged with the responsibility of networking all these five PRS systems for the purpose of creating a Distributed Passenger Reservation System by integrating all the separate systems and their databases. Thus CONCERT (Countrywide Network of Computerized Enquiry and Reservation Terminals) was born which fully replaced the existing system in 1999 enabling bookings from anywhere to anywhere from across any Railway Reservation Office in the country. CRIS had been mandated to continue with the existing hardware platform which were DEC VAX/VMS systems.

The architecture of the initial PRS system did not allow for such a federated database based transaction processing model. Fundamental changes to the architecture and consequently the design and programming of the systems had to be done to achieve this. A Transaction Processing Monitor - DEC's Reliable Transaction Router (RTR) - was introduced to support highly reliable distributed transaction processing. I was the Chief Project Manager and the Architect of this project and was involved from the very beginning of the project - right from the time of conceiving the project based on IR's requirement, through to its development, trial runs and then complete implementation across all the server locations. I built a new team from scratch to get this implemented. The entire project was developed in-house by the team at CRIS which was managed by me.

2. When and why was IIT Delhi approached for involvement in this project? How did Prof Maheshwari get involved in the process?

One of the key requirements of the system was to have a highly scalable, reliable and high performance filing system. Since the system was meant to withstand high level of concurrent usage, especially during peak hours, the need was to choose the right file system which would enable a high throughput of transactions. Those were the days when the computing power and storage access speeds were not as powerful as they are today. Also, RDBMS technology had not matured enough to be in the reckoning for heavy concurrent usage OLTP systems. The OLTP systems of those days developed by IBM, UNISYS etc. were all based on proprietary flat file systems.

To address the challenge of building a very efficient file system Professor Maheshwari, whom CRIS had engaged as a technology consultant, suggested an appropriate in-memory index based Dynamic Hashing (DHASH) based file system. He guided the team through the development of this high performance file system which was used to store millions of Passenger Name Records (PNRs). Professor Maheshwari's inputs, provided as part of comprehensive and critical reviews of the system, were absolutely critical to the realization of this ambitious and critical project for Indian Railways.

My association with Prof Maheshwari was since my B.Tech days and subsequently he was also my guide for my M.Tech project. He was very excited about this project and I remember discussing with him the challenges we had been facing for which he would provide very useful advice. Subsequently he agreed for an engagement with CRIS as a consultant for the new generation PRS.

3. Can you share what the experience of working with academia partners was like for you on such a major project? Any interesting stories/anecdotes to relate?

Professor Maheshwari's involvement and guidance was absolutely critical to the success of this project. His guidance and inputs from contemporary academic research in highly critical areas clearly made the difference between the success and failure of the project. The choice of the right algorithm, the critical reviews to ensure that the implementation is such that it mirrors the key concept behind the hashing algorithm were all ensured by Professor Maheshwari. The eventual proof of this is in the fact that the system is still being used close to two decades after it was developed and has survived not only exponential growth in its usage but also fast paced evolution in computing related technologies.

4. How do you view the changed technology scenario since the time this project was undertaken? What kind of role do you see for IITD in it?

I am convinced that Academia has a highly critical role in the industry. Most of the innovation that we see today is based on academic research which has been in the works in previous decades. Today we are seeing the industry exploiting concepts that are based out of academic research performed in institutions such as IITD - whether it is in the area of Artificial Intelligence, Cognitive Computing, Semantic Computing, Machine Learning, Robotics, Bio-informatics, nano technologies, Virtual reality etc.

In the hyper competitive business landscape of today's globalized world, the reality today is that Companies need innovation not just to thrive but even to survive. Hence corporate houses today are keen to tie up with academic institutions of excellence such as IITD for joint research and consultancy.

Copyright © 2017 Department of Computer Science and Engineering. All Rights Reserved.