Replication Techniques for Availability

Abstract The chapter studies how to provide clients with access to a replicated object that is logically indistinguishable from accessing a single yet highly available object. We study this problem under two different models. In the first, we assume that failures can be detected accurately. In the second we drop this assumption, making the model more realistic but also significantly more challenging. Under the first model, we present the primary-backup and chain replication techniques. Under the second model, we present techniques based on voting. We conclude with a discussion on reconfiguration.

• Type: Journal
• Year: 2010
• Where: REPLICATION
• Authors: Robbert van Renesse, Rachid Guerraoui
• Keywords: Chain Replication Techniques, Computational Biology, Model, Replicated Object, REPLICATION 2010