Huawei Cloud Storage helps CERN discover God particle

Founded in 1954, CERN (European from for Nuclear Research), headquartered in Geneva, Switzerland, is the birthplace of www and the largest particle physics study organization in the World, It provides particle accelerators and other infrastructure for the study of high energy Physics for international cooperative experiments, such as the source of the universe, antimatter and so on. CERN has the world's largest LHC (Large Hadron Collider) system and the only 0-level data center in the LHC computing grid.

Storage challenges from massive data

The LHC's particle collision experiment will produce a huge amount of scientific data, which collects and transmits experimental data through 150 million sensors distributed in the annular tunnels. At least 20P of new mass data storage requirements each year are the first major challenges to the scalability of the IT system at CERN; However, in such a short collision time, the data generated per second to storage up to 6GB, which has a higher throughput capacity of the storage system; Each time CERN invested a large amount of human and material resources to carry out experiments, the Hard-won experimental data appear to be invaluable, storage systems need to match the high reliability; the flexibility of the experimental data to be geographically shared is an important basis for the collaborative work of CERN's research institutions around the world.

As a result, CERN's storage systems face daunting challenges in terms of scalability, read-write performance, reliability, and cost of construction, forcing CERN to urgently find new storage systems and solutions to improve the status quo and meet the more complex and demanding scientific needs of the future.

Huawei's UDs system is easy to handle

At the beginning of 2012, Huawei provided 0.8PB of UDs mass storage devices to CERN OpenLab, the only authoritative organization for validating IT Infrastructure and solutions at CERN, to explore, test and validate the capabilities of the UDs in a massive data scenario with CERN.

CERN believes that the key indicators of storage systems are performance, reliability, and construction costs; higher performance or higher reliability requires higher input costs. Based on a deep understanding of massive data services and storage systems, CERN has tested and validated the Huawei UDs mass storage System in several ways:

Reliability: In nearly a year of testing, the UDs system did not lose any data and the test results were evaluated by CERN as "outstanding reliability measured:100%".

Compatibility: The UDS system completes docking and interconnect operations with other systems (such as the Open stack system) through the S3 interface and successfully passes system compatibility verification.

Functional: CERN has been highly recognized and well received by CERN for the basic features of the UDS storage and the unique MDC (multiple data Center Unified resource pool), one-click Fast deployment, and instant-free maintenance.

Performance: One of the highlights of the distributed architecture of the UDS system is that as the number of nodes increases, the performance of the system can continue to improve. After testing at CERN for a long time, the throughput of the UDs has reached the limit of the bandwidth limit, and the performance test concludes "fully achieved expected configured".

Tco:uds system uses the industry's unique arm architecture, and in the premise of ensuring high reliability to take the lead in support of consumer-level hard disk, significantly reduce capex; the machine's high integration design, ultra-low power consumption and free real-time maintenance features are better to help customers significantly reduce opex.

In addition, the multiple data center features unique to the UDs ensure that data across different regions flows across a unified pool of resources, enabling CERN to collaborate better with its global research organization.