Update Propagation Strategies to Improve Freshness of Data in Lazy Master Schemes

Many distributed database applications need to replicate data to improve data availability and query response time. The two-phase-commit protocol guarantees mutual consistency of replicated data but does not provide good performance. Lazy replication has been used as an alternative solution. In this case, mutual consistency is relaxed and the concept of freshness is used to measure the deviation between replica copies. In this paper we present a framework for lazy replication and focus on a special replication scheme called lazy master. In this scheme the common update propagation strategy used is deferred update propagation and works as follows: changes on a primary copy are first commited at the master node, afterwards the secondary copy is updated in a separate transaction at the slave node. We propose strategies based on what we call immediate update propagation. With immediate update propagation, updates to a primary copy are propagated towards a secondary copy as soon as they occur at the master node without waiting for the commitment of the update transaction. We study the behavior of these strategies and show that immediate update propagation may improve freshness with respect to the deferred approach.