Building clouds to boost academic research in Japan

Image courtesy www.perspecsys.com, Flickr (CC BY-SA 2.0).

The following article was originally published inSUCRE Cloud Source Magazine under the headline 'What’s new? Japan’s research into cloud computing for academia'. Read more articles like this in the latest issue, here.

Across Japan, universities and government departments are accelerating their research into cloud computing for academic use. Their research aims not only to share cloud-computing resources between universities, thus reducing the cost of research, but also to share know-how on the development, operation, and management of cloud-computing systems for further research.

Japanese universities in action

There are many activities in universities in Japan for academic cloud computing. In November 2011, Hokkaido University launched the Hokkaido University Academic Cloud, based on CloudStack, an open-source cloud-computing software platform. This service has since been used by universities all over Japan. It provides more than 40 teraFLOPS of computing power.

In April 2014, a cloud storage service was added, offering two petabytes of distributed cloud storage. The service now provides more than 2,000 virtual servers and supports a huge volume of jobs, making it one of the most powerful academic cloud-computing infrastructures in Japan.

Many Japanese universities are also working to connect their cloud-computing infrastructures, aiming to build an inter-cloud environment that will give researchers access to more powerful resources, thus accelerating research.

SINET tops more than 700 universities and research institutions

The academic network, the Science Information Network (SINET), is one such inter-university offering, and universities across Japan are working towards connecting each university’s own cloud servers to this network.

SINET now has more than 700 universities and research institutions, and many of the researchers in those organizations have different research requirements. They use the academic cloud service for a wide range of research and analysis projects, from big-data analysis to artificial intelligence research. As some research areas necessitate the use of high-performance computing, some reserchers need access to multi-core virtual machines.

Government support

Many ministries within the Japanese government support research into cloud computing:

• In 2013, the Ministry of Education, Culture, Sports, Science and Technology funded research into the use of academic clouds for big-data research, gathering data from universities and other educational organizations, including Kyushu University and the Japanese National Institute of Informatics.
• Also in 2013, the Ministry of Economy, Trade and Industry funded research into the development of high-quality cloud technologies with high availability and low energy requirements. To this end, software-defined networking (SDN) has been picked up as one of the key technologies, along with a focus on the definition and standardization of application programming interfaces (APIs).
• Since 2009, the Ministry of Internal Affairs and Communications has funded the development of interoperability in cloud computing, with an eye to its use in disaster recovery.

Interoperability initiatives

Another focus for Japan’s cloud research is inter-cloud standardization. The Global Inter-Cloud Technology Forum, launched in 2009, provided its Intercloud Interface Specification Draft in 2012. In addition to this publication, inter-cloud standardization has also been discussed since 2013 in Study Group 13 of the International Telecommunication Union, Telecommunication Standardization Sector (ITU-T).

Japan’s contributions

Japanese universities and the Japanese government have advanced research and development of cloud computing to better utilize distributed computing resources. These efforts benefit not only cloud-computing technology, but also the many research areas reliant on huge computing resources.

There is much still to do: many cloud-related technologies, such as distributed processing, interoperability, SDN, machine learning, and artificial intelligence, as well as standardization activities, will benefit from further research, and advancing this will be one of Japan’s most impressive contributions.