作者: Ozalp Babaoglu , Alberto Bartoli
DOI:
关键词: Partition (database) 、 The Internet 、 Communication in small groups 、 Sequential consistency 、 Server 、 Distributed computing 、 Suspect 、 Exploit 、 Computer network 、 Load balancing (computing) 、 Computer science
摘要: Group communication is a middleware technology that may simplify development of highly available applications through replication. In this paper we propose replicated service implementation based on partitionable group communication. Unlike its primary-partition counterpart, system never blocks an application because network failures have caused it to partition. Modern Internet-based services with workloads are typically biased in favor read-only requests can exploit property and remain multiple partitions despite the presence state. The proposed offers several interesting features including sequential consistency, load balancing, ``atmost-once’’ semantics for updates unreliable between clients servers. our implementation, simply resubmit any server if they suspect original request been lost due broken TCP connection. possibility recover from by having reissue servers without incurring risk executions increasingly important requirement many services. makes use novel abstractions be their own right.