Re:Safety and liveness: Eventual consistency is not safe
by cperciva on Tuesday March 27, 2012 @09:21PM
(#3764251)
As Dylan16807 said, the key property in "eventually known consistency" is that when partitions don't happen you'll eventually get a "yes, that write has propagated everywhere" answer. You can do this with gossiping updates and two clock vectors (vector #1 is "I have seen all the writes node X did up to this time", and vector #2 is "node X says it has seen everything done by any node prior to this time").
While you have partitions, you'll get a "maybe" back, because you can't distinguish between the cases of "a write hasn't propagated everywhere" and "it has propagated everywhere but I haven't received an ACK for it"; that's unavoidable, but doesn't mean that this is not useful anyway.
Re:Safety and liveness: Eventual consistency is not safe
by Dylan16807 on Tuesday March 27, 2012 @07:10PM
(#3763922)
I think you missed this line.
> "yes" is guaranteed to eventually be returned after some bounded period of non-partition
Re:Safety and liveness: Eventual consistency is not safe
by pbailis on Tuesday March 27, 2012 @08:26PM
(#3764115)
If you can bound partition durations, you can definitely make stronger guarantees.
If you can model your network delays, you can use some modeling like our work on PBS (Probabilistically Bounded Staleness) to predict staleness: http://pbs.cs.berkeley.edu/#demo
For any given T, if I partition each of your nodes for T+1 seconds, you won't be able to guarantee convergence--your nodes won't communicate. Am I missing something?