It’s all action in Lapland…
Author(s): Peter Bye, Posted on December 19th, 2011
I understand that it’s all systems go at SantaPrise in Lapland. The nerve centre of the operation is SantaSystems, the IT arm of the company. The CIO, Elf Tietokone, spoke to me the other day, and described their IT systems and how they cope with the annual demand. Here’s what I heard.
Annual demand is of course the key phrase. In a seasonal business such as this, the load builds up for a couple of months, reaching a peak about now. Having systems that can respond to these peaks, while maintaining performance, reliability and security, is absolutely crucial.
The entire operation is guided by SPA, the service-oriented SantaPrise Architecture. The core applications at SantaSystems run in ClearPath systems, all of which are metered. The main client-facing application is order taking, which is implemented in Dorado systems in an XTC configuration. Metering, says Elf, provides huge amounts of power when required while keeping the annual cost to a minimum. And, when added to general high reliability of ClearPath systems, XTC provides exceptional resilience.
Clients are world-wide, using a variety of channels, including Web browsers, Web services, smart phones, tablets, text messages, voice input and emails. A particularly interesting channel is the letter-up-chimney network, which has proved quite tricky to implement. SantaPrise solved the problem by installing text scanners at the top of chimneys, connected via satellite to a central point, where the messages are wrapped into Web services.
The other side of the operation is manufacturing. ClearPath Libra systems run a factory management application which schedules manufacturing. Materials and stock control are also handled by Libra systems, allowing production to meet the demand with excellent consistency.
Three other systems support the organisation. Scheduling the optimum routing for delivery is a huge task, entrusted to a super-computer delivering over 1.5 Santaflops. A Dorado-based RDMS database contains information on good and naughty children – who gets presents and who doesn’t – while an MCP system manages the finances.
Following the architecture defined in SPA, the systems are interlinked on a high-performance SSB (SantaPrise Service Bus). Various technologies are used to communicate between the systems, depending on requirements. They include Web services, message queuing and file transfer.
Elf stresses that the data centre infrastructure has been designed to avoid loss of service no matter what happens. DR sites are remote, with fully-automated DR implemented using Operations Sentinel, OpCon/XPS, BCA and SRDF. The off-peak season allows time to enhance automation and carry out regular DR testing. The combination of automation and testing minimises any chance that the service will not be available when needed.
Of course, SantaPrise does not work in isolation. It partners with a number of other commercial and government organisations. Transport, for example has been outsourced to the Ren Deere Company, which gets information on routing and packing from SantaPrise. Other partners include banks, payment systems and suppliers of materials, as well as government organisations such as customs and air traffic control. The SSB supports many technologies for external communication, including Web services, file transfer, and other interfaces using Enterprise Output Manager – emails, faxes, PDF files and more – formatted to meet the partner’s needs.
Finally, Elf says that corporate social responsibility is taken very seriously. The location in Lapland has enabled the design of very efficient data centres, requiring minimal cooling power. And staff welfare is a priority, with a significant investment in elf and safety.
Happy Christmas and all good wishes for 2012!