Contents
Editorial
Tribune
7i program
Business cases
Experts voice
Solutions
At a glance
Quick poll
Events
PDF version
 

Subscribe to
Bull Direct:


 

Archives
N°27  |  June   2008
Business cases

• OnAir (France)
• Ministry of Finance (Egypt)
• Cardiff University (United Kingdom)

GSIT (France)
Orange (Poland)

GSIT: exemplary evolution of a highly complex system

Interview with Sylvain Omnès, Managing Director of GSIT (Groupement pour un Système Interbancaire de Télécompensation), the governing body of the French inter-bank clearing system

In France, all interbank exchanges are automated via of a single, unique system: the interbank clearing system SIT (Système Interbancaire de Télécompensation). Handling 12.4 billion operations every year worth some €5,200 billion, SIT is Europe’s largest channel for bulk processing of inter-bank transfers. Its security and operating quality are excellent with 99.99% service delivery.
In 2007, Bull led the project to update this highly complex system within an extremely tight schedule, to the full satisfaction of its operating body, GSIT.
What led you to undertake such a significant evolution of SIT in 2007?
Our on-going imperative is to maintain service continuity while delivering optimum performance and security. The overall availability of the Accounting Center – the heart of our system for regulating and authenticating inter-bank transfers – had been running at an excellent level, close to 100% for the last six years. However, technical limitations came to light in 2007, and in light of the STET project – which would extend the field of responsibility for all the European players in this area – we really had to update the SIT infrastructure. Our initial evaluations identified that Bull’s NovaScale 7005 platforms were capable of hosting the SIT Accounting Center application and delivering the necessary levels of operational continuity to meet our service commitments to the banking community.

How did you go about this?
A prototype created in conjunction with Bull in October 2006 had already established the feasibility of the operation, especially in terms of performance and security. Our system handles between 50 and 60 million transfer operations every day, and the Accounting Center itself has to process a minimum of 15 to 20 messages a second. Taking these constraints into account, the systems update had to take place over a nine-month period: a serious challenge! With Bull’s help, we optimized the implementation schedule and put in place a joint Bull-GSIT project team, who totally shared the objectives. Similarly, a reward/penalty scheme linked to the commitment to results was set up, which worked extremely effectively.

How did you qualify the Accounting Center architecture?
This is an extremely complex architecture, requiring the highest levels of availability, security and performance. The functional application architecture features its own disaster recovery center that can be activated from cold. The operational architecture is completely secure by virtue of the resilience and redundancy of all hardware and software components, to the extent that any individual component can be removed while the system is live. This is just one of the highly original features of the system architecture designed by Bull, which can also detect system incidents automatically, in real time. The second key strength of the system is related to security. We decided to switch over to IP, and as a result the Accounting Center is treated like a ’militarized zone’, and is totally secure against attacks of any kind. The third major strong point of the architecture was about the automatic mechanisms which handle production and monitor the system itself and the application, which involved a huge commitment from the teams involved.

What lessons can be learned from a project of this scale?
This project was managed in an exemplary way. The core of the system was changed with total transparency as far as the banks were concerned. The new Accounting Center platforms went into production during the course of Q4 2007, just nine months after the project started. It was a faultless implementation, but we achieve all the pre-requisites for success:

  • Using a prototype to establish the foundations of the system in terms of high availability, security and performance
  • Optimizing the infrastructure, following exhaustive system testing
  • A joint team consisting of the best experts from Bull and GSIT, with shared objectives and commitments to results
  • Finally, a high level of involvement from senior management at both Bull and GSIT.

I also believe that this was an exemplary systems evolution project, given the complexity of our system and the challenges it had to meet.

Contact  |  Site map  |  Legal  |  Privacy