首页 > 科技 > 计算机 > 网络与互联网

数据中心智能调度关键技术与应用

作译者：田文洪,徐敏贤,薛瑞尼

出版时间：2025-09 千字数：595 版次：01-01 页数：368

开本：16开装帧： I S B N ：9787121511394

换版：

纸质书定价：￥118.0

库存:无

分享到：

共有图书评论0条【查看评论摘要】

看了又看

{{data.bookName}}

￥{{data.price}}

内容简介

前言

上架建议

作者简介

编辑推荐

音视频专区

第1章 数据中心概述 ······································································.1
1.1 数据中心简介 ····················································································.2
1.1.1 什么是数据中心 ·········································································.2
1.1.2 数据中心的需求和挑战 ································································.4
1.2 云计算数据中心资源调度需求分析 ·························································.4
1.2.1 技术需求 ··················································································.4
1.2.2 技术目标 ··················································································.5
1.3 云计算数据中心资源调度研究进展 ·························································.5
1.4 云计算数据中心资源调度方案分析 ·························································.6
1.4.1 Google解决方案 ········································································.6
1.4.2 Amazon解决方案 ·······································································.7
1.4.3 IBM解决方案 ············································································.8
1.4.4 HP解决方案··············································································10
1.4.5 VMware解决方案 ·······································································10
1.4.6 阿里云解决方案 ·········································································12
1.4.7 华为云解决方案 ·········································································14
1.4.8 其他厂家解决方案 ······································································15
1.5 云计算数据中心资源调度标准进展 ·························································17
1.6 云资源管理调度关键技术及研究热点 ······················································18
本章小结 ·································································································20
思考题·····································································································21
参考文献 ·································································································21
第2章 云计算概述 ········································································.25
2.1 云计算的发展背景 ··············································································26
2.2 云计算是集大成者 ··············································································28
2.2.1 并行计算 ················································································.28
2.2.2 网格计算 ················································································.29
2.2.3 效用计算 ················································································.29
2.2.4 普适计算 ················································································.30
2.2.5 软件即服务 ·············································································.30
2.2.6 虚拟化技术 ·············································································.31
2.3 云计算的驱动因素 ············································································.31
2.3.1 云计算的发展现状和趋势 ···························································.33
2.3.2 云计算应用初步分类 ·································································.35
2.4 云计算产业链中的不同角色 ································································.36
2.5 云计算的主要特征和技术挑战 ·····························································.37
2.5.1 云计算的主要特征 ····································································.37
2.5.2 挑战性问题 ·············································································.38
本章小结 ································································································.46
思考题 ···································································································.46
参考文献 ································································································.46
第3章 大数据处理 ········································································.51
3.1 大数据的发展背景及定义 ···································································.52
3.2 大数据问题 ······················································································.55
3.2.1 速度方面的问题 ·······································································.55
3.2.2 种类及架构问题 ·······································································.56
3.2.3 体量及灵活性问题 ····································································.56
3.2.4 成本问题 ················································································.57
3.2.5 价值挖掘问题 ··········································································.57
3.2.6 存储及安全问题 ·······································································.58
3.2.7 互联互通与数据共享问题 ···························································.59
3.3 大数据与云计算的辩证关系 ································································.60
3.4 大数据技术 ······················································································.61
3.4.1 基础架构支持 ··········································································.62
3.4.2 数据采集 ················································································.64
3.4.3 数据存储 ················································································.65
3.4.4 数据计算 ················································································.68
3.4.5 数据展现与交互 ·······································································.73
本章小结 ································································································.75
思考题 ···································································································.76
参考文献 ································································································.76
第4章 人工智能平台的资源调度概述 ················································.78
4.1 引言 ·································································································79
4.2 深度学习的分布式并行训练系统架构 ······················································79
4.3 深度学习的分布式并行策略 ··································································81
4.3.1 深度学习的基础概念 ···································································82
4.3.2 分布式并行训练算法 ···································································82
4.3.3 研究现状分析 ············································································85
4.4 分布式并行训练的时效分析 ··································································91
本章小结 ·································································································96
思考题·····································································································96
参考文献 ·································································································97
第5章 基于深度学习的云服务负载预测方法 ·······································.99
5.1 引言 ······························································································.100
5.2 相关工作 ························································································.101
5.2.1 基于回归方法的云服务负载预测方法···········································.101
5.2.2 基于学习的云服务负载预测方法 ·················································.102
5.2.3 讨论分析 ···············································································.103
5.3 系统模型 ························································································.104
5.4 esDNN：基于高效监督学习的深度神经网络 ··········································.106
5.4.1 多元时间序列预测的滑动窗口 ····················································.106
5.4.2 esDNN算法 ···········································································.109
5.5 性能测试 ························································································.112
5.5.1 数据集和环境配置 ···································································.112
5.5.2 与基于无监督学习方法的比较 ····················································.113
5.5.3 与基于神经网络方法的比较 ·······················································.114
5.5.4 与其他方面的比较 ···································································.118
本章小结 ······························································································.119
思考题··································································································.120
参考文献 ······························································································.120
第6章 云应用程序和可再生能源的自适应管理方法 ····························.123
6.1 引言 ······························································································.124
6.2 相关工作 ························································································.125
6.2.1 DVFS和虚拟机整合·································································.125
6.2.2 Brownout ···············································································.126
6.2.3 数据中心冷却系统的整体管理 ····················································.126
6.2.4 可再生能源 ············································································.126
6.3 系统模型 ························································································.127
6.4 问题建模 ························································································.129
6.4.1 功率消耗 ···············································································.129
6.4.2 工作负载模型 ·········································································.130
6.4.3 优化目标 ···············································································.131
6.5 根据可再生资源进行调度决策 ····························································.132
6.5.1 Green-Aware 调度算法 ·····························································.132
6.5.2 交互式工作负载的Brownout算法 ···············································.133
6.5.3 批处理工作负载的延迟算法 ·······················································.134
6.5.4 主机调度 ···············································································.135
6.5.5 可再生能源预测 ······································································.136
6.6 原型系统的实现 ··············································································.137
6.7 性能评估 ························································································.139
6.7.1 环境设置 ···············································································.140
6.7.2 工作负载 ···············································································.140
6.7.3 应用 ·····················································································.141
6.7.4 结果 ·····················································································.141
本章小结 ·······························································································.145
思考题 ··································································································.145
参考文献 ·······························································································.146
第7章 云计算环境下基于虚拟机整合的自适应节能调度 ······················.149
7.1 绪论 ······························································································.150
7.2 虚拟机整合技术 ··············································································.150
7.3 相关研究工作 ··················································································.152
7.4 问题定义 ························································································.153
7.5 数据中心能耗模型 ···········································································.154
7.5.1 服务器功耗模型 ······································································.154
7.5.2 服务器能耗模型 ······································································.155
7.5.3 云数据中心总能耗模型 ·····························································.156
7.5.4 数据中心节能调度下限 ·····························································.156
7.6 SAVE算法描述 ···············································································.158
7.6.1 概述 ·····················································································.158
7.6.2 分配算法 ···············································································.159
7.6.3 迁移算法 ···············································································.161
7.7 实验验证与分析 ··············································································.165
7.7.1 实验准备 ···············································································.165
7.7.2 数据准备 ···············································································.167
7.7.3 基线方法 ···············································································.167
7.7.4 结果分析 ···············································································.169
本章小结 ······························································································.175
思考题··································································································.175
参考文献 ······························································································.175
第8章 MapReduce模型中数据倾斜问题的算法 ·································.177
8.1 绪论 ······························································································.178
8.1.1 背景及意义 ············································································.178
8.1.2 研究现状 ···············································································.179
8.1.3 研究内容 ···············································································.180
8.2 数据倾斜相关理论研究 ·····································································.181
8.2.1 数据倾斜 ···············································································.181
8.2.2 算法介绍 ···············································································.183
8.2.3 算法综合对比 ·········································································.190
8.3 多任务数据倾斜调度算法设计 ····························································.193
8.3.1 问题描述与建模 ······································································.193
8.3.2 Revised Johnson1954算法（RJA） ··············································.195
8.3.3 离线多任务调度算法 ································································.199
8.3.4 在线多任务调度算法 ································································.201
8.4 单任务数据倾斜算法设计 ··································································.203
8.4.1 YarnTune概述 ·········································································.203
8.4.2 检测数据倾斜 ·········································································.206
8.4.3 YarnTune核心功能 ··································································.208
8.5 系统测试和分析 ··············································································.212
8.5.1 软硬件测试环境 ······································································.212
8.5.2 多任务数据倾斜测试 ································································.212
8.5.3 单任务数据倾斜测试 ································································.217
本章小结 ······························································································.220
思考题··································································································.221
参考文献 ······························································································.221
第9章 Spark中的数据均衡分配算法研究 ·········································.223
9.1 Spark设计思想 ················································································.224
9.1.1 Spark概述 ·············································································.224
9.1.2 Spark计算模型 ·······································································.225
9.1.3 Spark整体架构 ·······································································.226
9.2 Spark数据存储体系 ··········································································.227
9.2.1 存储整体架构 ·········································································.227
9.2.2 数据写入过程 ·········································································.228
9.2.3 数据读取过程 ·········································································.229
9.3 Spark Shuffle分析 ············································································.230
9.3.1 Shuffle概述 ············································································.230
9.3.2 Shuffle写操作 ·········································································.231
9.3.3 Shuffle读操作 ·········································································.232
9.4 Spark分区方法 ················································································.233
9.4.1 HashPartition分区 ····································································.233
9.4.2 RangePartition分区 ··································································.234
9.5 问题描述与建模 ··············································································.235
9.5.1 相关定义 ···············································································.235
9.5.2 问题建模 ···············································································.236
9.6 数据均衡分配算法整体设计 ·······························································.237
9.6.1 抽样算法 ···············································································.238
9.6.2 数据均衡分区算法 ···································································.240
9.6.3 权重调整算法 ·········································································.242
9.6.4 任务分配算法 ·········································································.245
9.7 算法复杂度分析 ··············································································.246
9.8 MRFair概述 ···················································································.246
9.8.1 MRFair的目标与特征 ·······························································.246
9.8.2 MRFair系统架构 ·····································································.247
9.8.3 MRFair数据均衡分配示例 ·························································.248
9.9 MRFair倾斜检测时机及算法 ······························································.249
9.9.1 MRFair倾斜检测时机 ·······························································.249
9.9.2 MRFair倾斜检测算法 ·······························································.250
9.10 MRFair数据重新分配时机及算法 ······················································.250
9.10.1 MRFair数据重新分配时机 ·······················································.250
9.10.2 MRFair数据重新分配算法 ·······················································.251
9.11 MRFair核心模块 ············································································.253
9.12 系统测试环境 ················································································.255
9.12.1 软硬件测试环境 ·····································································.255
9.12.2 测试数据 ··············································································.256
9.12.3 对比算法或系统 ·····································································.257
9.12.4 评价指标 ··············································································.257
9.13 Reduce Partition数据均衡分配算法测试 ··············································.257
9.13.1 WordCount基准测试 ·······························································.257
9.13.2 Sort基准测试 ········································································.260
9.14 MRFair数据均衡分配算法测试 ·························································.263
9.14.1 WordCount基准测试 ·······························································.263
9.14.2 Sort基准测试 ········································································.265
本章小结 ······························································································.266
思考题··································································································.267
参考文献 ······························································································.267
第10章 深度学习框架TensorFlow的高效分布式并行算法研究 ·············.269
10.1 分布式并行算法的背景及意义 ···························································.270
10.1.1 问题背景 ··············································································.270
10.1.2 研究意义 ··············································································.271
10.2 研究现状及内容 ·············································································.272
10.2.1 研究现状 ··············································································.272
10.2.2 研究内容 ··············································································.272
10.3 深度学习理论研究 ··········································································.273
10.3.1 大数据与云计算 ····································································.273
10.3.2 机器学习 ··············································································.274
10.3.3 深度学习 ··············································································.275
10.4 TensorFlow深度学习框架研究 ··························································.277
10.4.1 TensorFlow系统架构 ······························································.277
10.4.2 TensorFlow数据流图 ······························································.280
10.4.3 TensorFlow会话管理 ······························································.281
10.4.4 TensorFlow分布式执行 ···························································.282
10.4.5 TensorFlow数据输入 ······························································.283
10.5 TensorFlow分布式架构分析 ·····························································.285
10.5.1 TensorFlow远程调用 ······························································.285
10.5.2 现有TensorFlow分布式模型 ····················································.286
10.6 优化算法设计与实现 ·······································································.289
10.6.1 数据并行上的优化 ·································································.289
10.6.2 模型并行上的优化 ·································································.297
10.7 实验环境配置 ················································································.304
10.7.1 硬件环境 ··············································································.304
10.7.2 软件环境 ··············································································.304
10.7.3 实验对象 ··············································································.305
10.7.4 实验数据 ··············································································.305
10.8 实验结果展示与分析 ·······································································.307
10.8.1 数据并行算法测试··································································.307
10.8.2 模型并行算法测试··································································.312
本章小结 ·······························································································.314
思考题 ··································································································.314
参考文献 ·······························································································.314
第11章 基于深度强化学习和模仿学习的任务完工时间优化调度 ···········.317
11.1 任务调度 ······················································································.318
11.2 相关研究 ······················································································.321
11.3 云资源调度问题定义 ·······································································.322
11.4 调度方案 ······················································································.327
11.4.1 DeepRM_Online介绍 ······························································.327
11.4.2 强化学习模型 ········································································.328
11.4.3 深度强化学习训练算法····························································.330
11.5 算法分析 ······················································································.333
11.6 实验分析与验证 ·············································································.337
11.6.1 实验准备 ··············································································.337
11.6.2 数据准备 ··············································································.337
11.6.3 基线算法 ··············································································.338
11.6.4 结果分析 ··············································································.339
本章小结 ·······························································································.341
思考题 ··································································································.341
参考文献 ·······························································································.342
第12章 总结与展望 ······································································.345
12.1 绿色节能数据中心的综合解决方案·····················································.346
12.2 多数据中心（多调度域）的调度策略和算法动态可选择 ·························.348
12.3 支持深度学习模型的分布式并行调度 ·················································.349
12.4 从基础资源调度拓展到应用任务调度 ·················································.350
参考文献 ·······························································································.351

序 言
云计算已成为推动IT行业包括人工智能领域发展的重要力量之一。云计算提供基于订购的基础设施、平台和应用访问服务，通常被称为基础设施即服务（Infrastructure as a Service，IaaS）、平台即服务（Platform as a Service，PaaS）和软件即服务（Software as a Service，SaaS）。这些服务将计算和应用托管的成本降低了数个数量级，但在智能调度，以无缝、可伸缩、可靠的方式开发和交付等方面仍存在重大挑战。
其中一个具有挑战性的问题是实现高效的云资源优化调度，而本书是少数几部专注于IaaS级别资源优化调度的专著之一。多数数据中心目前只实现了简单的调度策略，存在资源利用效率低、处理速度慢，以及难以适配外部负载和内部应用等问题。资源优化调度主要解决优化建模、优化资源分配和动态实时迁移问题，以提升效率与质量。本书聚焦解决这些具有代表性的问题，考虑了多维资源（CPU、存储、网络等）的负载均衡和能效等特性，而不仅是静态的参数预设。为了实现高性能、节能和降低成本的目标，云数据中心需要在动态环境中高效地处理物理资源和虚拟资源融合的问题。本书介绍前沿性研究成果，包括使用深度学习方法对云计算负载进行精准预测、利用深度强化学习方法进行任务调度等。
本书旨在促进科学、工业、商业和消费者应用等领域的云数据中心的计算资源高效管理和优化调度技术，推动云计算相关领域的发展。本书还提供了关于云计算资源调度和管理的最新技术综述。我强烈推荐本书作为系统架构师、实践者、开发人员、新研究人员、研究生及高年级本科生的学习与参考资料。
郑纬民
中国工程院院士、清华大学教授
2024.4.25
前 言
书中内容需发挥，莫为技艰便后退。
纵使智能难跟随，入得深处令人追。
——《书中留客》 田文洪
“经过精细规划的优化理论设计的实践比随意性或一般性实施在性能、节能及提高运营利润等方面可体现高出多个量级的效果，并不断接近或达到最优化结果。”
数据中心：不少研究者预言“国家间将来的核心竞争在数据中心”。数据中心是容纳计算设备资源的集中之地，同时负责计算设备的能源提供和空调维护等。数据中心可以单独建设，也可以置于其他建筑之内，还可以是分布在不同地理位置的多个系统。云资源汇聚在一起，通过多租户模式服务多个消费者。在物理上，资源以分布式的共享方式存在，但最终在逻辑上以单一整体的形式呈现给用户。资源种类很多，分类角度也不一样，本书所涉及的资源主要包括以下几类。
（1）物理服务器：构成数据中心的物理计算设备，每个物理服务器可以提供多个虚拟机，每个物理服务器可以由多个CPU、内存、硬盘、网卡等构成。
（2）物理集群：由多个物理服务器、必要的网络和存储设施构成的物理服务器组。
（3）虚拟机：通过虚拟化软件在物理服务器上生成的虚拟计算平台，可以由多个虚拟化CPU、硬盘、网卡等构成。
（4）虚拟集群：由多个虚拟机、必要的网络和存储设施构成的虚拟机组。
（5）共享存储：为数据中心的计算资源提供大容量存储，可以被所有设备和用户共享。
以上数据中心资源在本书中统称为基础资源，简称云资源。
云计算：云计算是一种计算模型和服务模式，它将计算任务分布在大量计算机构成的不同数据中心，使各种应用系统能够根据需要获取计算能力、存储空间和信息服务。提供资源的网络或数据中心被称为“云”。业界研究者将云计算列为水、电、气、油之外的第五种公用资源（The Fifth Utility）。继个人计算机、互联网变革之后，云计算被看作第三次IT浪潮，是世界各国战略性新兴产业的重要组成部分，它将带来生活、生产方式和商业模式的深刻改变，已成为当前全社会关注的热点。
云计算目前已经广泛应用于网络搜索、科学计算、虚拟环境、能源和生物信息等领域的日常业务和创新性探索。云计算应用将以政府、电信、教育、医疗、金融、石油石化和电力等行业为重点，在中国市场逐步被越来越多的企业和机构采用，市场规模也将越来越大。
云资源优化管理调度技术是云计算应用的核心，是云计算得以大规模应用和提高系统性能、兼顾节能减排及运营成本等方面的关键技术。先进的动态资源调度管理，对于提高学校、政府、研究机构和企业计算资源的利用效率，节约能源、提高资源共享和降低运营成本具有极大的意义，值得深入系统地学习和研究。
大数据：指无法在一定时间内用常规软件工具对其内容进行抓取、管理和处理的数据集合，需要新型的数据处理模式。大数据具备“5V”特征：Volume（数据体量巨大）、Variety（数据格式多样）、Velocity（数据产生速度快，流量大）、Value（数据的总体价值在增大，单位价值密度却在降低）、Veracity（数据真实准确）。传统的信息技术架构，存在扩展性差、容错性差、性能低、安装部署及维护困难等诸多瓶颈，已无法处理大数据问题，需要以现代云计算的手段和技术来解决大数据问题。云计算与大数据是相辅相成、辩证统一的关系，云计算、物联网技术的广泛应用是我们的愿景，而大数据的爆发则是发展中遇到的棘手问题。大数据技术的突破不仅能解决现实困难，还会促使云计算、物联网技术真正落地并深入推广和应用。大数据处理关键技术及大数据处理平台的优化调度技术，包括批量任务和单个任务的处理性能都需要深入研究以提升效率。
人工智能技术：近些年，随着人工智能技术的不断发展，数据、计算和存储等内容的需求量也在不断上升。国家在人工智能发展方面的重大战略，推动了国内数据中心、智能计算中心等人工智能基础设施这类“新基建”的发展。根据《新一代人工智能发展规划》的相关内容，到2025年，我国人工智能基础理论能够实现一些重大突破，部分技术也要求能够达到世界领先水平。这些非常具有前瞻性的规划，为人工智能的研究及应用提供了巨大的空间，也会进一步推动专业化的针对人工智能研究及应用的数据中心和算力平台的发展。有别于普通的数据中心，人工智能平台需要的智能计算中心将针对人工智能技术相关的核心技术进行优化，使得人工智能的相关模型、算法能够具有最高的运行效率；同时，人工智能平台需要能够根据实际情况，动态地调整全局资源使计算中心的运维开销最小化。结合以上内容，可以发现相对于传统的服务器机房、数据中心等设施而言，智能计算中心需要一个强大的人工智能平台，该平台不仅要能管理计算中心的服务器，还要能动态地对存储及算力等内容进行实时动态的调配。因此，人工智能平台本身也需要具有一定的“智能”调度，根据算法研究的实际情况，动态地分析当前物理环境的全局状态，结合存储、算力、网络等各种信息对相关模型的训练、推理进行自动编排，充分地利用平台分布式算力对当前人工智能应用中最核心的操作进行提速，同时最大化自身算力资源利用率。在深度学习场景下，人工智能平台需要实现平台资源的合理调度，在优化资源使用效率的同时缩短大规模深度学习模型的训练和推理时间，并保证模型的精度与泛化能力。
自2023年以来，大语言模型的热度持续升高，成为国内外各大AI科技巨头激烈竞争的热门赛道之一。例如，OpenAI的ChatGPT、百度的文心一言，这些应用已经逐渐融进人们的生活，在一定程度上改变了人们的生活方式，提高了工作效率。但由于这类模型不仅参数规模巨大，而且对计算资源、存储效率和数据吞吐量的要求极高，因此，人工智能平台作为大语言模型的部署和调度引擎，在提高资源使用率、缩短训练和推理时间，减小服务响应延时产生的收益方面甚至高于模型本身在神经网络结构上的优化。
资源管理调度是将资源从资源提供方分配给用户的一个过程。对于资源过载（需求大于系统容量），以及需求与系统容量随着时间的推移而动态变化的事实，优化资源调度管理，可以动态地重新分配资源，以便更高效地使用可用资源。业界一般将调度分为作业级调度和设施级调度。作业级调度是指针对具体的运行程序，系统将作业具体分配到哪些资源上运行的问题。例如，一些独立的需要较多计算资源和较长运行时间的程序、高性能并行处理程序，或者大数据处理和人工智能分布式并行训练程序，往往需要较大规模的高性能计算资源（如云计算）才能很快完成。设施级调度主要指将底层资源作为一种基础设施即服务（IaaS）提供给用户，用户依据实际情况使用这些资源。例如，数据中心的物理服务器（含CPU、内 存、网络带宽等）、虚拟机（含虚拟化的CPU、内存、网络带宽等）及虚拟集群都属于底层基础资源。
本书侧重于设施级（也称应用级）调度，兼顾作业级调度（如MapReduce多任务调 度），并介绍资源动态优化管理与调度。如果把数据中心作为一个整体与人体比较，资源优化管理调度类似人的大脑，是最为核心的功能之一，具有极大的理论意义和实用价值。另外，资源监控类似人的眼睛，资源部署类似人的四肢。目前的数据中心管理与调度大多仅实现了简单的初级功能，还有很多问题亟须深入系统解决。针对基础资源优化管理调度，本书主要回答以下三大基本问题。
（1）优化目标问题：将所需的虚拟机（或虚拟集群）配置在某个数据中心物理服务器（或物理集群）的具体优化函数（目标）是什么？如何建立高效的数学理论优化模型？这涉及宏观的管理策略制定。
（2）优化分配问题：具体在哪个数据中心的哪个物理服务器（或物理集群）上分配所需的虚拟机（或虚拟集群）以实现优化目标？满足的具体前提和限制条件有哪些？针对不同的优化目标，业界优化实践方案有哪些？
（3）实践性能问题：针对理论优化目标与模型，业界优化管理调度方案的性能对比效果如何？如何选择适合不同目标的具体方法？
“没有理论的实践是盲目的，没有实践的理论是苍白的。”从优化和长远的角度来看，云计算的建设需要坚实的理论与实践经验的结合。经过多年数据中心建设的实践和理论探索，我们在本书中提出并总结相关领域的优化理论模型和实践方法，在解决基本问题的同时，实现动态调度而不是静态预设全部相关参数，兼顾资源（CPU、存储、网络等）的性能、热耗、利用率等特性，本书称为资源的动态综合调度技术问题。
数据中心需要处理物理资源和虚拟资源动态结合的优化管理调度这一新问题，以实现高性能、节能减排及降低投资成本等目标。当前，数据中心的资源调度不少仍然沿袭传统或简单的方法，难以满足以上目标的精细化和动态变化的要求。数据中心调度面临的挑战性问题包括：在考虑配置动态可调虚拟机的分配和迁移及物理机综合性能，兼顾CPU、存储、网络等多资源因素的情况下，如何解决用户需求不一致和资源规格不一致造成的系统性能不平 衡、能耗效率低下及成本过高等问题。
本书围绕数据中心IT基础资源智能优化调度关键技术问题，介绍数据中心、云计算和大数据及深度学习计算平台的发展背景与资源调度的挑战性问题，对资源智能优化调度领域内的主要挑战性问题进行了深入分析和探索，特别是对云负载预测、绿色节能调度、大数据平台优化调度、深度学习框架的优化调度等关键内容进行了深入分析与讨论，希望为读者深入了解相关知识提供一些借鉴。
本书各章之间的关系紧密围绕数据中心资源智能调度关键技术与应用展开。本书主要内容包括数据中心概述（第1章）、云计算概述（第2章）、大数据处理（第3章）、人工智能平台的资源调度概述（第4章）、基于深度学习的云服务负载预测方法（第5章）、云应用程序和可再生能源的自适应管理方法（第6章）、云计算环境下基于虚拟机整合的自适应节能调度（第7章）、MapReduce模型中数据倾斜问题的算法（第8章）、Spark中的数据均衡分配算法研究（第9章）、深度学习框架TensorFlow的高效分布式并行算法研究（第10章）、基于深度强化学习和模仿学习的任务完工时间优化调度（第11章），以及总结与展望（第12章）。本书可作为高年级本科生和研究生的教材，也可作为相关专业研究人员的参考资料。
本书是编者对以上内容大量理论知识与实践经验的积累结果，书中难免存在不妥之处，欢迎读者指正，请发邮件至CloudSched@gmail.com。
编著者
2024年7月

查看更多 > 图书评论

暂无评论

发表图书评论

评论标题：
评论内容：
验证码：: 看不清楚
点击刷新
: 您还没有登录，请登录后再评论。
点击“登录”

购买过本书的顾客还买过

{{data.bookName}}

￥{{data.price}}