Post-Partition Reconciliation Protocols for Maintaining Consistency
2006 (English)Conference paper (Refereed)
This paper addresses design exploration for protocols thatare employed in systems with availability-consistency tradeoﬀs. Distributed data is modelled as states of objects replicated across a network, and whose updates require satisfaction of integrity constraints over multiple objects. Upondetection of a partition, such a network will continue toprovide delivery of services in parallel partitions; but onlyfor updates with non-critical integrity constraints. Once thedegraded mode ends, the parallel network partitions are reconciled to arrive at one partition.Using a formal treatment of the reconciliation process,three algorithms are proposed and studied in terms of theirinﬂuence on service outage duration. The longer the reconciliation time, the lower is system availability; since theinterval in which no services are provided is longer. However, the reconciliation time in turn is aﬀected by the time toconstruct the post-partition system state. The shorter theconstruction time the higher is the number of updates thattook place in the degraded mode but that will not be takenup in the reconciled partition. This will lead to a longerinterval for rejecting/redoing these operations and therebyincrease reconciliation time.
Place, publisher, year, edition, pages
ACM , 2006.
reconciliation protocol, availability, partition, consistency, trade-oﬀ
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-72607DOI: 10.1145/1141277.1141439OAI: oai:DiVA.org:liu-72607DiVA: diva2:460373
The 21st ACM/SIGAPP symposium on Applied computing