CERN ditches Microsoft for open source

Read time 3min 30sec

One of the world’s most prestigious and best-known research facilities is 'taking back control' of its computing activities by moving away from Microsoft's commercial software to open source.

This was announced last week by CERN, Switzerland-based European Organisation for Nuclear Research in a blog which, for the first time, provided details of its Microsoft Alternatives project (MAlt).

CERN is perhaps best known for its 2012 discovery of the Higgs Boson 'God particle', and its Large Hadron Collider (LHC). It has often been featured in movies and popular literature including in Dan Brown’s mystery thriller 'Angels and Demons'; and in the popular TV series 'The Big Bang Theory'.

According to the company’s system architect Emmanuel Ormancey, CERN had been using Microsoft software for some 20 years and received favourable licence conditions, thanks to Microsoft’s recognition of it as an academic, non-profit or research institution. 

However, in March this year, Microsoft revoked the research organisation’s academic status and replaced its existing contract with a new one based on user numbers. This would effectively increase CERN’s licence costs by more than a factor of 10.

“Given the collaborative nature of CERN and its wide community, a high number of licenses are required to deliver services to everyone, and when traditional business models on a per-user basis are applied, the costs per product can be huge and become unaffordable in the long term,” Ormancey wrote.

Muted response from Microsoft

Microsoft’s response to CERN’s announcement was muted. The Seattle-based software giant, which has been honing its overtures to the open source community of late, reportedly said only that: “We are committed to a consistent application of eligibility requirements for our different licensing options. When we find exceptions to this policy during our ongoing audit processes for compliance, we work directly with customers to ensure their transition to new licensing programs is as seamless as possible.”

Clearly this was not sufficient for CERN which acknowledged that it had negotiated with Microsoft to introduce the increase gradually over a period of 10 years in order to give the organisation time to adapt. Nevertheless, CERN did not believe these costs could be supported in the long term.

While the Microsoft Alternatives project is ambitious, it’s also a unique opportunity for CERN to demonstrate that building core services can be done without vendor and data lock-in.

The organisation had clearly anticipated this situation and so had launched MAlt a year ago to investigate the migration from Microsoft (and other) commercial software to open source “so as to minimise CERN’s exposure to the risks of unstainable commercial conditions.”

Its goal is to find solutions that will deliver the same service to every category of CERN personnel; avoid vendor lock-in to decrease risk and dependency on commercial software vendors; keep CERN’s hands on its own data; and address common use-cases.

Ormancey noted that CERN was not the only research institution to be faced with this dilemma and believed that the steps being taken by CERN would provide a roadmap for others to follow.

MAlt is billed as a multi-year effort, and it is entering a new phase with the first migrations currently taking place.

One of the first changes involves a pilot mail service for the IT department, prior to its roll out to the entire organisation. At the same time, CERN will be moving away from Skype for Business (and analogue phones) to a softphone telephony pilot.

Other systems and products will follow.

“While the Microsoft Alternatives project is ambitious, it’s also a unique opportunity for CERN to demonstrate that building core services can be done without vendor and data lock-in, that the next generation of services can be tailored to the community’s needs, and finally that CERN can inspire its partners by collaborating around a new range of products,” Ormancey concluded.

See also