3 Ways to Achieve Crash Consistency


Read how to keep data safe in the presence of crashes.

“Even well-managed data centers suffer from occasional power-loss events,” says Ram Alagappan in a recent article for ACM Queue, which means ensuring crash consistency across data storage systems is crucial.

“A storage system is deemed crash-consistent if it can recover the persistent data it stores to a meaningful state after crashes,” Alagappan says, but achieving this state can be quite challenging.

This article explores three ways (through three research papers) to help: “finding and fixing crash-consistency bugs; developing new abstractions that ease crash consistency; and exploiting new hardware.”

Read more at ACM Queue.


Related content

comments powered by Disqus