GitHub goes under the ice: 21 TB of open-source hidden in the arctic repository
As you may remember, Microsoft’s GitHub platform planned to create a secure archive of open source in the Arctic ice. More precisely, in a mine specially equipped for storage on the island of Spitsbergen, in permafrost conditions. The snapshot for the archive was taken on February 2, 2020, but due to the COVID-19 pandemic, the archive only reached storage this month.
As reported on the GitHub blog, a copy of the most important or related open source code was made on February 2 (almost everything, as GitHub, claims, but this is unlikely). “The GitHub Arctic Code Vault snapshot of 02/02/2020 covers all active public GitHub repositories, in addition to essential but dormant repositories. “
The snapshot includes each repository with commits between the announcement on GitHub on November 13, 2019, and February 2, 2020, each repository with a minimum of 1 star and commits one year before the snapshot (from 02/03/2019 to 02/02/2020), as well as each repository with 250 or more stars. But in practice, even the code without stars received the right to get into the Arctic archive, which makes it difficult to fully assess the rules for choosing the right to storage. Let’s add that GitHub users, whose code got into the Arctic archive, received labels with the entry “Arctic Code Vault Contributor”.
In total, 21 TB of data were selected for the archive, which was recorded on special tapes in Piql reels. This is a silver halide polyester film. Each reel has a little more than a kilometer of the film (1066 meters). A total of 186 such coils were recorded, and all of them in nine large boxes were delivered to the storage on the island of Svalbard in Norway. The pandemic prevented the GitHub team from personally escorting the archive to storage at a depth of several hundred meters, but, according to the company, the archive was successfully delivered and placed in the shah on July 8.
The Piql company, which owns the Arctic storage, guarantees the safety of data on the tape for 500 years. Artificial aging of the tape showed that the data from the tape can be counted even after 1000 years of storage. It remains to be hoped that by that time the means of reading such archives will remain on Earth, otherwise anything can happen. To save is one thing, and to understand why it is necessary and what to do with it is another.
