MIT researchers have created a system that can never lose data. Well, the data will always be recoverable. The team of six will showcase the system that is mathematically guaranteed not to lose data during crashes in October at the ACM Symposium on Operating Systems Principles.
In an operating system (OS) the file system is the part that records & writes data onto a disk. If the computer crashes while writing data, the data can become corrupt and to everyone’s despair, unreadable. In modern day the file system the MIT gang have conjured up may be a bit slow, but formal verification could make it a lot easier to generate efficient, reliable file systems.
One of the hardest parts of this goliath task was making sure the system can be recovered from any point. There are so many places and processes that could make a system crash, so this set of wizards are looking into every instruction and disk operation thinkable. Even in well-tested systems bugs will be found and they will keep on getting found, because it is just that hard to do.
The achievement is found in the file systems final code and not just the schematics. It wasn’t all magic though, the researchers had help from a tool called proof assistant.
Authors of this system are Haogang Chen; Daniel Ziegler; Tej Chajed; Adam Chlipala; M. Frans Kaashoek, and Nickolai Zeldovich
Kaashoek said “No one had done it, it’s not like you could look up a paper that says, ‘This is the way to do it.’ But now you can read our paper and presumably do it a lot faster.
We find this piece research fantastic and we are very excited for the system to be presented in October. Data backup is key to a business’s longevity and security. Even with this new file system we can’t stress how important it is to store your data in a safe and secure place, preferably offsite.
Thing of the past? It already is with Datacentreplus
