A general scalable and elastic matching service for content-based publish/subscribe systems

摘要

Characterized by the emergence of a large number of live content, the emergency applications have receivedrnincreasing attention in recent years. Providing a general and scalable event, matching service can preciselyrnnotify users latest information that they are interested in. However, because the live content arrival raternmay churn significantly in a short time and subscriptions with various patterns tend to be skewed, it isrnchallenging to increase the generality, scalability, and elasticity of the matching process.We propose a novelrnparallel event matching service based on the cloud computing environment, called GSEM, to satisfy thesernrequirements. GSEM first presents a two-hop framework and a general subscription pattern to handle variousrnpatterns of subscriptions. To provide scalable matching service, a hybrid content space partition scheme isrnproposed to divide large skewed subscriptions into multiple small clusters managed by a group of parallelrnservers. To adapt to the sudden change of event arrival rate, GSEM elastically adjusts the scale of servers andrnrebalances their workloads through a performance-aware detection technique. A prototype deployment onrnthe OpenStack platform shows that GSEM achieves scalable matching throughput with the growth of servers,rnelastic service capacity with the change of event arrival rate, and significantly outperforms the existing cloudrnbased systems in various workloads.