Planetary nebulae are dying stars that gain their beauty from shedding gaseous layers. Discovered in 1824 and resembling the human eye, the Helix Nebula is an iconic image in the world of astronomy. Nearly 200 years later, CERN has created the Helix Nebula initiative in partnership with leading IT providers and Europe’s biggest research institutes with the aim of establishing a science cloud to support large-scale research projects.
Planetary nebulae are dying stars that gain their beauty from shedding gaseous layers. Discovered in 1824 and resembling the human eye, the Helix Nebula is an iconic image in the world of astronomy. Nearly 200 years later, CERN has created the Helix Nebula initiative in partnership with leading IT providers and Europe’s biggest research institutes with the aim of establishing a science cloud to support large-scale research projects.
Cosmic achievements
Known for its commitment to unraveling the mysteries of the universe, CERN has added many strings of astounding achievements to its bow: In 1973, it discovered neutral current. Then ten years later, CERN’s UA1 and UA2 experiments found the W and Z bosons. In 2012, the ATLAS and CMS experiments on the Large Hadron Collider (LHC) discovered the now-famous Higgs boson particle — an elementary particle that completes the standard model of particle physics. Today, CERN continues to operate the world’s largest particle-physics lab, in search of exciting new discoveries that will help to explain the 96% of the known universe not covered by the standard model.
While most of the Organization’s achievements seem intangible and distant despite capturing the public’s imagination, it’s worth noting that Tim Berners-Lee established the World Wide Web at CERN in 1989, starting the online revolution that’s now stronger than ever.
A corner of the CERN’s lab. © CERN
Weighty projects like these demand equally weighty technology, computing and storage solutions.
CERN runs one of the world’s largest private OpenStack clouds. Located in Switzerland and Hungary, two data centers house 13,000 servers and 33,000 virtual machines, providing computing and storage capabilities for many European research institutes. However, future planned upgrades to the LHC and the experiments mean that the computing capacity required to analyze the data produced is likely to increase by a factor of 50-100 by 2026.
CERN is evaluating different models for increasing its computing and data-storage capacity, from federating multiple data centers under a common management and access model, to procuring resources from commercial providers, to building new computing centers on CERN’s premises. All models present different technological challenges.
Since 2012, CERN has played a leading role in a series of European projects under the ‘Helix Nebula’ banner. This initiative is now working to establish a European Open Science Cloud, which would enable large research laboratories to boost their computing resources seamlessly through hybrid computing models.
Challenge 1: Capacity and balance
The complexity of the LHC — as well as its planned schedule of upgrades — means that computing requirements can vary significantly over time. Hybrid computing models are therefore of significant interest to CERN: in-house computing resources can be rapidly bolstered by resources from commercial cloud providers at peak times of need.
Challenge 2: Rapid obsolescence
The rapid iteration of new technology is part of life in 2017. The planned obsolescence of smartphones, for example, leaves users behind if they don’t upgrade to models that offer the latest features. In a similar vein, greater use of cloud at large research organizations like CERN offers the potential advantage of making it easier to keep up with resource-upgrade cycles.
Challenge 3: Uniquely high requirements
Research institutes are different from common enterprises. Public clouds used for research must feature the highest standards when it comes to security, inoperability between institutions, and performance. Such clouds would have to meet European security standards, support OpenStack and open APIs, and provide data-sharing mechanisms.
Through the Helix Nebula initiative, CERN and other research institutes have been working to find suitable cloud computing platforms that, by combining public and commercial cloud service providers, are able to meet these needs.
Team effort
Led by CERN, a buyers group comprising the following research institutes forged a public-private partnership to search for a solution.
European Institute of Bioinformatics (EMBL)Italian Institute of Nuclear Physics (INFN)
German Institute for Electronic Synchrotron Accelerator (DESY)
France Center for Scientific Research (CNRS)
Karlsruhe Institute of Technology, Germany (KIT)
Spanish Institute of High Energy Physics (IFAE)
European Institute for Synchrotron Radiation Devices (ESRF)
UK Council for Science and Technology Facilities (STFC)
Netherlands National Super Computing and Electronic Technology Research Institute (SURFsara)
Design and testing
The Helix Nebula initiative recently concluded Phase 1: Solution Design and Pre-Commercial Procurement. In this phase, partner research institutions requested potential solution design and tests from prospective providers. In November 2016, CERN selected four consortia of prospective providers — from many bids — to design a solution. In April 2017, three solutions were selected to proceed to Phase 2: Prototyping. Deutsche Telekom and Huawei’s Open Telekom Cloud is a contender in this phase.
Solving challenges
So far, Open Telekom Cloud, powered by Huawei, has enjoyed a number of successes in the Helix Nebula initiative.
Helix Nebula project schematic diagram. © CERN
Open for all
Research institutes’ public clouds need to be open source. Open Telekom Cloud solution fully supports open-source requirements, with open APIs available to participating research bodies.
Synchronized sharing
Open Telekom Cloud provides a One Data solution that enables research institutes to create global access to data and then upload, retrieve, and share resources through a single sign-on account.
One Data allows users to upload data to different locations that is then seamlessly synchronized. For example, an institute could deploy an application in the public Open Telekom Cloud while storing its data in its private cloud. When applications are launched, all data would be synced across the public and private clouds, greatly enhancing transmission speeds and efficiency.
High-speed interconnection
During a Phase 2 test, Open Telekom Cloud supported high-speed interconnection between sites using a 10Gbit/s network connected to GÉANT, the European academic network. In the next phase, Open Telekom Cloud will be able to significantly increase speed to support 40Gbit/s network bandwidth.
High performance computing
Research institutes are at the forefront of innovation and science; it is important that providers can craft a suitable computing solution. In response to the high computing demands of the Helix Nebula partners, Open Telekom Cloud features a High Performance Computing (HPC) solution.
Security and capacity
Open Telekom Cloud has established an IaaS hybrid cloud ecosystem that works with CERN’s private cloud and meets European data security standards. Partner research institutes will be able to store highly confidential information, such as genetic data, in their private clouds while migrating other sharable data to public clouds. This enables the institutes to both free up private cloud capacity and accelerate data invocation.
Certification
Open Telekom Cloud is eduGAIN-certified, which is a mechanism for establishing trust between identity providers and service providers.
Invaluable experience
Together with Deutsche Telekom, Huawei has successfully designed, tested, and provided public cloud services for the Helix Nebula initiative.
The experience Huawei has gained with Open Telekom Cloud in the context of Helix Nebula will prove invaluable for other future projects in the research domain. Large scientific institutes can serve as role models for commercial enterprises embarking on digital-transformation journeys, providing a roadmap for overcoming technological challenges and driving future-looking innovations.
