Designing multi-level quorum schemes for highly replicated data

The authors present and analyze multi-level quorum schemes for maintaining the consistency of replicated data in the presence of concurrency and failures in a large distributed environment. The multi-level method operates on a logical hierarchy of replicas and applies well known flat voting algorithms in a layered fashion. They show how the number of hierarchy levels, the number of logical entities per level and the voting algorithms used on each level affect the costs and the degree of availability associated with a wide range of multi-level quorum schemes in order to design the most suitable multi-level strategy for a given application scenario. The analytical results obtained are confirmed by performance measurements in a simulated network