A Japanese university lost 77TB of scientific research data due to a backup error

Data backup is a mandatory and extremely important process for any organization or business in today’s digital age. For educational institutions, scientific research, data synchronization and backup is even more important. However, just planning or giving a reasonable data backup route is not enough, how to make the backup process happen correctly and safely is also a matter of concern.

Kyoto University, one of the leading scientific research institutions in Japan, has just encountered a terrible “catastrophe” when it lost most of its archives due to an unexpected malfunction in the supercomputer system. The Hewlett-Packard computer is responsible for backing up school data.

Specifically, a system crash unexpectedly occurred between December 14 and 16, 2021, causing 77 TB of Kyoto University data to be accidentally erased during the backup process. normally they still do it periodically. This amount of data is equivalent to about 34 million document files, coming from 14 different research groups. All have been removed from the system.

Preliminary investigation results show that the data of at least 4 research groups will be completely unrecoverable. This is a heavy loss and extremely regrettable for not only Kyoto University, but also the field of science in general, as it is one of the largest research institutions in Asia, regularly receiving grants. large grants to serve the purpose of scientific research. Every year, Kyoto University makes great contributions in a variety of core fields such as biology, pharmacology, immunology, materials science and physics.

Currently, the backup process has been paused for in-depth testing and evaluation processes to prevent data loss from happening again:

“While we have addressed this issue, it is likely that we have lost up to 100TB of data and will continue to investigate further the extent of the impact it caused.”

The Kyoto University data storage process is currently performed by the Hewlett Packard Cray supercomputer and the DataDirect ExaScaler storage system. These are all systems that many different organizations and research institutions use in their data management processes. Therefore, this incident is not only the story of Kyoto University.

Another aspect to mention is that the cost of operating these supercomputer systems for data backup purposes is also not cheap. Running a supercomputer alone costs hundreds of dollars an hour.

However, like other machines and electronic devices, supercomputers can completely encounter problems during operation. While this doesn’t happen often, the damage it causes can often be huge.


Related Posts

Leave a Reply

Your email address will not be published.