Groupama Asset Management is the the 8th asset management company in France. A subsidiary of Groupama, employing over 200 persons. Present in two different market, Groupama has managed around 52.6 billion euros in 2003.

The Informatics continuity plan (ICP) is aimed to:

• The protection of the power supply
• The increasing of the electric and air cooling capacity per sq. meter
• The increasing of the machine occupation density
• The on-demand extension of the of the available surface, according to the current needs
• Protection from social conflicts and from terrorism
• Availability and decrease of the optical fibers costs
• Access to several telecommunications operators

To solve all the problem faced, two different options could be considered:

• Total replacement of the existing infrastructure:
  Systems installed on a new site with the latest technology
  Low migration risk
  Important financial investment
  No activity interruption in the relocation period
  Potential temporary costs relating to the MAN
• Relocation of the infrastructure:
  The existing infrastructures are relocated during the moving period.
  Risks for the data and the equipment is higher but the cost of the operation is lower than the first solution.
  The relocation period would necessarily imply the a service interruption and the implementation period would be longer

After study, an intermediary solution appeared worth considering: breaking up the architecture. With this solution: the risk incurred during the transfer is reduced, no service interruption is to be foreseen, the implementation period even though greater than if in the replacement solution is shorter than that of the relocation, and the MAN cost remains recurrent.

Use of clusters and recourse to a classical Development, Reception, Integration and Production scheme was not sufficient as the building had become the architecture's SPOF. A time limit of 4 hours was set for the restart of the client's activities, while integrity of its data had to be secured and the transfer could be done without interrupting the service.

The solution consisted in the creation of two computing sites connected in real-time and to create a single virtual one with two remote premises. The objective was then to find, on the same LAN, SAN and WAN networks with a single administration, the same tools and technologies on the two sites, the use of a high-availability architecture so that the sole investments be made for missing equipment and the use of the integration, the receipt and development architecture to install the servers on both sites. The MAN then has to be the backbone of the project.