On Sun, Apr 24, 2011 at 6:26 PM, lingli deng <[email protected]> wrote: > Hi all, > > I am concerned about the specific leaving procedure for Chord > algorithm described in Chapter 9, which does not address the data > migration issue from the leaving peer to the newly responsible peer. > Is it assumed that the newly responsible peer is one of the backup > peers in Chord?
Yes, that's how Chord works. > Even if it is the case, there seems to be possibility for loss of data > if the leaving peer does not perform data backup right before its > exit. The leaving peer is supposed to replicate the data when it is stored, so in general loss should be minimized. If the client really cares, it can check the replicas to see if the store happens and if not re-store. > Would this be a problem? Well, it can happen, but there's no guarantee that data is never lost, it's just a "with low probability" anyway. IMO no fix is required here. Best, -Ekr > BR, > > Lingli Deng > _______________________________________________ > P2PSIP mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/p2psip > _______________________________________________ P2PSIP mailing list [email protected] https://www.ietf.org/mailman/listinfo/p2psip
