Current capacity planning practices based on heavy over-provisioning of power infrastructure hurt (i) the operational costs of data centers as well as (ii) the computational work they can support. We explore a combination of statistical multiplexing techniques to improve the utilization of the power hierarchy within a data center. At the highest level of the power hierarchy, we employ controlled underprovisioning and over-booking of power needs of hosted workloads. At the lower levels, we introduce the novel notion of soft fuses to flexibly distribute provisioned power among hosted workloads based on their needs. Our techniques are built upon a measurement-driven profiling and prediction framework to characterize key statistical properties of the power needs of hosted workloads and their aggregates. We characterize the gains in terms of the amount of computational work (CPU cycles) per provisioned unit of powerComputation per Provisioned Watt (CPW). Our technique is able to double the CPW offered by a Power Distribution Unit (PDU) running the e-commerce benchmark TPC-W compared to conventional provisioning practices. Over-booking the PDU by 10% based on tails of power profiles yields a further improvement of 20%. Reactive techniques implemented on our Xen VMM-based servers dynamically modulate CPU DVFS states to ensure power draw below the limits imposed by soft fuses. Finally, information captured in our profiles also provide ways of controlling application performance degradation despite overbooking. The 95 th percentile of TPC-W session response time only grew from 1.59 sec to 1.78 sec-a degradation of 12%.
With the ever increasing amount of traffic, scalability is probably the most important factor that differentiates several existing approaches to traffic classification. This paper focuses on payload-based classification and compares the results obtained through a "lightweight" traffic classification approach with the ones obtained with a "completely stateful" approach, demonstrating that the first approach, albeit less precise, is still appropriate for a large class of applications.
This paper presents a novel approach in traffic classification that is based on the identification of the service that generates the traffic. This method is, in some sense, orthogonal to current approaches and it can be used as an efficient complement to existing methods to reduce computation and memory requirements. Experimental results on real traffic confirm that this method is extremely effective and may improve considerably the accuracy of traffic classification, while it is suitable to a large number of applications.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.