基于区块链的毕业设计Cloud Computing Architecture Project – 云计算架构项目

本文提供基于区块链的毕业设计国外最新区块链项目源码下载,包括solidity,eth,fabric等blockchain区块链,基于区块链的毕业设计Cloud Computing Architecture Project – 云计算架构项目 是一篇很好的国外资料

Cloud Computing Architecture Project

This repository contains code, results and reports for the Cloud Computing Architecture course project at ETH Zurich. The project explores how to schedule latency-sensitive and batch applications in a cloud cluster.

  • Part 1: Analyze the sensitivity of latency sensitive cloud application (memcache) with various types of resource interference (CPU, L1 instruction and data cache, L2 cache, L3 cache and Memory Bandwidth interference).
  • Part 2: Analyze the sensitivity of batch cloud applications (PARSEC) to various types of resource interference (same as part 1) and parallelism.
  • Part 3: Design an interference-aware scheduling policy to run latency-critical and batch applications on shared cluster.
  • Part 4: Design an interference-aware scheduling policy to run the latency-critical applications on the same VM.

Results

  • The static scheduler for part 3 achieves an SLO of 2ms for 95th percentile latency at 30k QPS.
  • The dynamic scheduler for part 4 achieves an SLO of 2ms for 95th percentile latency for loads varying between 5k QPS and 100k QPS with 8 second intervals in between next waves.

Files and Folders

  • cloud_comp_project_parts1-4.pdf: This file contains the detailed project description and instructions for setting up the clusters required for the project.
  • interference: This folder contains the YAML files used induce resource contention in the clusters.
  • memcache: This folder contains the YAML files used to set up memcache pod for parts 1 & 3.
  • parsec-benchmarks: This folder contains the YAML files used to set up PARSEC pods for parts 2 & 3.
  • project_yaml_files: This folder contains the YAML files we used to start up the clusters used in parts 1-4.
  • reports: This folder contains our reports for the project. You can find the detailed explanations of memcache and PARSEC application profiling in report_part_1_2.pdf file. You can find detailed explanation of the design of our static and dynamic scheduling policies, and their respective performances in the report_part_3_4.pdf.
  • results: This folder contains the output logs of memcache and PARSEC executions for parts 1-4.
  • src: This folder contains the implementations of our scheduling policies. static_schedule.py contains the static scheduling policy for part 3. controller_final folder contains the implementation of the dynamic scheduling policy and controller for part 4.

Contributors

  • Julia Bazinska
  • Floris Westermann
  • Berke Egeli

云计算架构项目

此存储库包含苏黎世ETH云计算架构课程项目的代码、结果和报告。该项目探索如何在云集群中调度对延迟敏感的批处理应用程序

  • 第1部分:分析延迟敏感云应用(memcache)对各种资源干扰(CPU、一级指令和数据缓存、二级缓存、,三级缓存和内存带宽干扰)
  • 第2部分:分析批处理云应用程序(PARSEC)对各种类型的资源干扰(与第1部分相同)和并行性的敏感性
  • 第3部分:设计一个干扰感知调度策略,在共享集群上运行延迟关键型和批处理应用程序
  • 第4部分:设计一种干扰感知调度策略,以便在同一虚拟机上运行延迟关键型应用程序

结果

  • 第3部分的静态调度器在30k QPS的95%延迟下实现了2ms的SLO
  • 第4部分的动态调度器对于5k QPS和100k QPS之间的负载(在下一个波之间有8秒的间隔)实现了2ms的SLO(第95百分位延迟)

文件和文件夹

  • cloudu compu projectu parts1-4.pdf:此文件包含详细的项目描述和设置项目所需集群的说明
  • 干扰:此文件夹包含用于在集群中引发资源争用的YAML文件
  • memcache:此文件夹包含用于为parts 1&设置memcache pod的YAML文件;3.
  • parsec基准:此文件夹包含用于为第2部分和第2部分设置parsec pod的YAML文件;3.
  • 项目yamlu文件:此文件夹包含我们用于启动第1-4部分中使用的群集的yaml文件。
  • 报告:此文件夹包含我们的项目报告。您可以在reportu partu 1u 2.pdf文件中找到memcache和PARSEC应用程序评测的详细说明。您可以在报告u part u 3 u 4.pdf中找到有关静态和动态调度策略的设计及其各自性能的详细说明
  • 结果:此文件夹包含第1-4部分的memcache和PARSEC执行的输出日志。
  • src:此文件夹包含我们的调度策略的实现。static_schedule.py包含第3部分的静态调度策略。controller_final文件夹包含第4部分的动态调度策略和controller的实现。

贡献者

  • Julia Bazinska
  • Floris Westermann
  • Berke Egeli

部分转自网络,侵权联系删除区块链源码网

www.interchains.cc

https://www.interchains.cc/23036.html

区块链毕设网(www.interchains.cc)全网最靠谱的原创区块链毕设代做网站 部分资料来自网络,侵权联系删除! 最全最大的区块链源码站 ! QQ3039046426
区块链知识分享网, 以太坊dapp资源网, 区块链教程, fabric教程下载, 区块链书籍下载, 区块链资料下载, 区块链视频教程下载, 区块链基础教程, 区块链入门教程, 区块链资源 » 基于区块链的毕业设计Cloud Computing Architecture Project – 云计算架构项目

提供最优质的资源集合

立即查看 了解详情