Neptune: scalable replication management and programming support for cluster-based network services

Abstract

Previous research has addressed the scalability and availability issues associated with the construction of cluster-based network services. This paper studies the clustering of replicated services when the persistent service data is frequently updated. To this end we propose Neptune, an infrastructural middleware that provides a flexible interface to aggregate and replicate existing service modules. Neptune accommodates a variety of underlying storage mechanisms, maintains dynamic and location-transparent service mapping to isolate faulty modules and enforce replica consistency. Furthermore, it allows efficient use of a multi-level replica consistency model with staleness control at its highest level. This paper describes Neptune's overall architecture, data replication support, and the results of our performance evaluation. 1 Introduction High availability, incremental scalability, and manageability are some of the key challenges faced by designers of Internet-scale network services...