| Document Date: 2011-12-07 19:08:05
Open Document File Size: 167,85 KBShare Result on Facebook
City Supercomputing / Urbana / New York / Washington / DC / / Company Checkpoint / Amazon / S. Inc. / Google / Fast Long-Distance Networks / IEEE Journal / VMWare / Intel / LWN.net / / Country United States / / / Facility Consider Amazon’s Elastic Block Store / Eric Brewer University of California / / IndustryTerm datacenter applications / performance computing / multi-core operating systems / opportunity to rethink how operating systems / distributed multimedia applications / regular block device / Memory resource management / internet services / datacenter hardware / real-time community / resource management / software mechanisms / performance operating systems / consistent systems / lightweight solution / recent operating systems / parallel systems / multicore systems / large computing facilities / parallel applications / software stack / esx server / data processing frameworks / e.g. applications / lowlatency processing / data processing / cluster-management software / large media files / application-specific memory management / energy management / multi-lane high speed networking cards / energy efficiency / data-intensive and processing-intensive nodes / datacenter management software / computing / cloud computing environment / software interface / appropriate hardware / real-time systems / bulk processing / datacenter operating systems / datacenter computing / Legacy software / performance applications / parallel processing / server systems / heterogeneous hardware / application-level management / legacy applications / Cloud computing workloads / web servers / virtual processor / legacy systems / user-level network protocol / Individual applications / network packet processing / commodity operating systems / capability computing / similar customized systems / memory management / / NaturalFeature Orcas Island / / OperatingSystem VMs / BSD / Nemesis / Microsoft Windows / Linux / Multics / GNU / POSIX / / Organization UC Berkeley / USENIX Association / University of California / Berkeley / Eric Brewer University / IEEE Computer Society / / Person A. Vahdat / M. Conley / K. Klues / V / G. Porter / A. Pucher / R. Mysore / H. Madhyasha / Kevin Klues / David Zhu / / Position thread scheduler / Hadoop scheduler / kernel scheduler / General / application-level memory manager / author / application thread scheduler / pthread scheduler / WAY FORWARD / Akaros kernel scheduler / user/system administrator / second-level scheduler / node-local Dominant Resource Fairness scheduler / buffer manager / application memory manager / custom second-level scheduler / scheduler / simple scheduler / Porter / / ProgrammingLanguage php / DC / C++ / / ProvinceOrState New York / Illinois / / Technology FPGA / same networking/RPC protocol / RAM / TritonSort algorithm / Linux / API / user-level network protocol / operating system / Quality of Service / html / large-scale SMP processors / virtual memory / virtual machine / operating systems / Ethernet / sorting algorithm / http / virtual processor / paging / massively parallel processing / caching / network protocols / / URL http /
SocialTag |
Improving Per-Node Efficiency in the Datacenter with New OS Abstractions Barret Rhoden, Kevin Klues, David Zhu, Eric Brewer University of California, Berkeley Berkeley, CA {brho, klueska, yuzhu, brewer}@cs.berkeley.edu