数据中心:修订间差异
删除的内容 添加的内容
JimGrassroot(留言 | 贡献) →历史: 内容扩充 增加或调整参考来源 |
无编辑摘要 |
||
(未显示2个用户的21个中间版本) | |||
第10行:
[[File:CERN Server 03.jpg|thumb|[[歐洲核子研究組織]](CERN)的資料中心]]
'''数据中心'''({{lang-en|Data center}}),指用于安置[[电子计算机|计算机]]系统及相关部件的设施,例如[[电信]]和[[電腦數據存貯器|储存]]系统。一般它包含[[系统冗余|冗餘]]和备用[[電源供應]],冗余[[数据通信]]连接,环境控制(例如[[空调]]、[[灭火器]])和各种安全设备。大型数据中心消耗的电約與一个小城镇工业业务规模一样多。<ref name=NYT92212>{{cite news|title=Power, Pollution and the Internet|url=https://www.nytimes.com/2012/09/23/technology/data-centers-waste-vast-amounts-of-energy-belying-industry-image.html|access-date=2012-09-25|newspaper=The New York Times|date=September 22, 2012|author=James Glanz|archive-date=2019-05-16|archive-url=https://web.archive.org/web/20190516142830/https://www.nytimes.com/2012/09/23/technology/data-centers-waste-vast-amounts-of-energy-belying-industry-image.html|dead-url=no}}</ref> 据估计,2022年全球数据中心的耗电量为 240-340 太瓦时 ([[瓦特|TWH]]),约占全球电力需求的 1-1.3%。这不包括加密货币挖矿所用的能源,据估计,2022 年加密货币挖矿所用能源约为 110 [[瓦特|TWH]],占全球电力需求的0.4%。<ref>{{Cite web |title=Data centres & networks |url=https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks |access-date=2023-10-07 |website=IEA |language=en-GB |archive-date=2023-10-06 |archive-url=https://web.archive.org/web/20231006181048/https://www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks |dead-url=no }}</ref>
数据中心在规模、电力要求、冗余度和整体结构方面差异很大。用于划分数据中心类型的四个常见类别是现场数据中心 (onsite data centers)、主机代管设施 (colocation facilities)、超大规模数据中心hyperscale data centers、和边缘数据中心 (edge data centers)。<ref>{{Cite web |title=Types of Data Centers {{!}} How do you Choose the Right Data Center? |url=https://www.maysteel.com/blog/types-of-data-centers-how-do-you-choose-the-right-data-center/ |access-date=2023-10-07 |website=Maysteel Industries, LLC |archive-date=2023-06-01 |archive-url=https://web.archive.org/web/20230601152244/https://maysteel.com/blog/types-of-data-centers-how-do-you-choose-the-right-data-center |dead-url=no }}</ref>
== 历史 ==
第21行:
数据中心是全球协作的特定设备网络,用来在网络基础设施上传递、加速、展示、计算、存储数据信息。
在繁荣的微型计算机行业,尤其是在
[[File:Datacenter de ARSAT.jpg|thumb|ARSAT卫星公司的数据中心(2014年)]]
▲在繁荣的微型计算机行业,尤其是在20世纪80年代,计算机开始四处发展,在很多案例中很少有或没有操作需求。然而,随着信息技术(IT)的发展,操作开始变得复杂,公司渐渐地认识到需要控制IT资源。随着客户端-服务器计算的出现,在20世纪90年代,微型机(被称为servers)开始在旧计算机房间中寻找他们的位置。便宜网络设备的可用性,外加网络电缆的新标准,使得在公司内的一个房间中使用分层设计来放置[[伺服器|服务器]]的可能,在这个时刻被公认并不断流行开来。
1997年至2000年[[互联网泡沫]]期间,数据中心迎来了繁荣。<ref name=DataM>{{cite magazine|magazine=Datamation|author=Cynthia Harvey|date=July 10, 2017|url=https://www.datamation.com/data-center/what-is-data-center.html|title=Data Center|access-date=2024-07-28|archive-date=2022-12-16|archive-url=https://web.archive.org/web/20221216132242/https://www.datamation.com/data-center/what-is-data-center/|dead-url=no}}</ref><ref group="note">There was considerable construction of data centers during the early 2000s, in the [[Dot-com bubble|period of expanding dot-com businesses]].</ref>公司需要快速的網絡连接,以及不断的部署系统并出现在網絡上。安装这种设备对于许多小公司是不可行的。许多公司开始建设非常大的设施,称为互联网数据中心(IDC),<ref name=IDC2K.NYT>{{cite news|newspaper=[[The New York Times]]|url=https://www.nytimes.com/2000/05/14/realestate/commercial-property-engine-room-for-internet-combining-data-center-with-telco.html|title=Commercial Property/Engine Room for the Internet; Combining a Data Center With a 'Telco Hotel'|author=John Holusha|date=May 14, 2000|access-date=June 23, 2019|archive-date=2024-06-28|archive-url=https://web.archive.org/web/20240628163920/https://www.nytimes.com/2000/05/14/realestate/commercial-property-engine-room-for-internet-combining-data-center-with-telco.html|dead-url=no}}</ref> 其提供增强的功能,例如交叉备份:“如果贝尔大西洋公司的一条线路被切断,我们可以将它们转移到……以最大限度地减少停机时间。”
它提供了商业上的系统部署和操作的解决方案。新技术和实践被设计用来处理测量和如此巨大测量操作的操作需求。
从2007年起,数据中心设计、构建和运作是一个众所周知的学科。它的标准文档来自于可信任的专家组,如电讯产业联合会,详细说明数据中心设计的需求。众所周知数据中心可用性的操作韵律学能够用来评估商业中断的影响。有许多开发在操作实践中被完成,并且设计了友好环境的数据中心。
'''云数据中心''' (cloud data centers, CDCs) 这一术语已被使用。<ref>{{cite journal |author=H Yuan |title=Workload-Aware Request Routing in Cloud Data Center |journal=Journal of Systems Engineering and Electronics |doi=10.1109/JSEE.2015.00020 |s2cid=59487957 |doi-access=free}}</ref> 数据中心的建设和维护通常成本很高。<ref name=DataM/> 这些术语之间的区别越来越小,它们正被整合到'''数据中心'''这一术语中。<ref>{{cite web|website=[[The New York Times]]|url=https://bits.blogs.nytimes.com/2011/10/04/a-data-center-power-solution|title=A Data Center Power Solution|author=Quentin Hardy|date=October 4, 2011|access-date=2024-07-28|archive-date=2022-12-16|archive-url=https://web.archive.org/web/20221216132243/https://archive.nytimes.com/bits.blogs.nytimes.com/2011/10/04/a-data-center-power-solution/|dead-url=no}}</ref>
2010年代,全球数据中心市场稳步增长,后半段增长明显加快。据[[高德纳咨询公司]] (Gartner) 称,尽管[[2019冠状病毒病疫情]]带来了经济挑战,但2021年全球数据中心基础设施支出达到$2000亿美元,比2020年增长了 6%。<ref>{{cite news |title=Gartner Says Worldwide Data Center Infrastructure Spending to Grow 6% in 2021 |url=https://www.gartner.com/en/newsroom/press-releases/2020-10-07-gartner-says-worldwide-data-center-infrastructure-spending-to-grow-6-percent-in-2021 |access-date=6 July 2024 |publisher=Gartner, Inc. |archive-date=2024-07-06 |archive-url=https://web.archive.org/web/20240706031816/https://www.gartner.com/en/newsroom/press-releases/2020-10-07-gartner-says-worldwide-data-center-infrastructure-spending-to-grow-6-percent-in-2021 |dead-url=no }}</ref>
2010年代后期和2020年代初,[[人工智能]]和[[机器学习]]应用发生了重大转变,推动了全球对更强大、更高效的数据中心基础设施的需求。截至2021 年3月,全球数据生成量预计将从2020年的64.2 [[泽字节]] (ZB)增长到2025年的 180 ZB以上。<ref>{{cite web |last1=Petroc |first1=Taylor |title=Volume of data/information created, captured, copied, and consumed worldwide from 2010 to 2020, with forecasts from 2021 to 2025 (in zettabytes) |url=https://www.statista.com/statistics/871513/worldwide-data-created/ |website=Statista |access-date=6 July 2024 |archive-date=2024-08-30 |archive-url=https://web.archive.org/web/20240830140921/https://www.statista.com/statistics/871513/worldwide-data-created/ |dead-url=no }}</ref>
在最近的发展热潮中,美国已确立了数据中心基础设施的领先地位,截至2024年3月,美国拥有5,381个数据中心,是全球数据中心数量最多的国家。<ref>{{cite web |last1=Petroc |first1=Taylor |title=Leading countries by number of data centers as of March 2024 |url=https://www.statista.com/statistics/1228433/data-centers-worldwide-by-country/ |website=Statista |access-date=2024-07-28 |archive-date=2022-12-16 |archive-url=https://web.archive.org/web/20221216132254/https://www.statista.com/statistics/1228433/data-centers-worldwide-by-country/ |dead-url=no }}</ref> 据全球咨询公司[[麦肯锡公司]]称,到2030年,美国市场需求预计将从2022年的17 GW翻一番,达到35 GW (吉瓦)。<ref>{{cite web |title=Investing in the rising data center economy |url=https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/investing-in-the-rising-data-center-economy |website=McKinsey & Co. |access-date=6 July 2024 |archive-date=2024-08-30 |archive-url=https://web.archive.org/web/20240830162245/https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/investing-in-the-rising-data-center-economy |dead-url=no }}</ref>截至 2023 年,美国约占全球市场的40%。<ref>{{cite web |title=Investing in the rising data center economy |url=https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/investing-in-the-rising-data-center-economy |website=McKinsey & Co. |access-date=6 July 2024 |archive-date=2024-08-30 |archive-url=https://web.archive.org/web/20240830162245/https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/investing-in-the-rising-data-center-economy |dead-url=no }}</ref>
美国电力研究所 (EPRI) 于2024年5月发布的一项研究估计,到2030年,美国数据中心的电力消耗可能占该国发电量的4.6%至9.1%。<ref>{{cite web |title=Powering Intelligence: Analyzing Artificial Intelligence and Data Center Energy Consumption |url=https://www.epri.com/research/products/3002028905 |website=Electric Power Research Institute (EPRI) |access-date=6 July 2024}}</ref> 截至2023 年,美国约80%的数据中心负载集中在15个州,其中[[弗吉尼亚州]]和[[德克萨斯州]]最多。<ref>{{cite web |title=Powering Intelligence: Analyzing Artificial Intelligence and Data Center Energy Consumption |url=https://www.epri.com/research/products/3002028905 |website=Electric Power Research Institute (EPRI) |access-date=6 July 2024 |archive-date=2024-08-22 |archive-url=https://web.archive.org/web/20240822171131/https://www.epri.com/research/products/3002028905 |dead-url=no }}</ref>
== 现代数据中心的需求 ==
[[File:Datacenter-telecom.jpg|thumb|left|数据中心部分区域的电信设备机架]]
[[File:Equipment-of-modern-conference-hall.jpg|缩略图|机架式服务器构成的机房]]
现代化和数据中心转型可提高性能和[[電效率]]。<ref name="mspmentor.net">{{cite web|url=http://www.mspmentor.net/2011/08/17/hp-updates-data-transformation-solutions/|title=Mukhar, Nicholas. "HP Updates Data Center Transformation Solutions," August 17, 2011|access-date=September 9, 2011|archive-date=August 12, 2012|archive-url=https://web.archive.org/web/20120812184028/http://www.mspmentor.net/2011/08/17/hp-updates-data-transformation-solutions/|url-status=dead}}</ref>
电信行业协会<ref>{{Cite web |title=Telecommunications Industry Association 942 Certifications & Ratings |url=https://tiaonline.org/products-and-services/tia942certification/tia-942-certifications-ratings/ |website=TIA Online |language=en-US |access-date=2022-10-02 |archive-date=2022-10-02 |archive-url=https://web.archive.org/web/20221002035027/https://tiaonline.org/products-and-services/tia942certification/tia-942-certifications-ratings/ |dead-url=no }}</ref>制定了数据中心电信基础设施标准,规定了数据中心和计算机机房基础设施(包括单用户企业数据中心和多用户网络数据中心)的最低要求,本文档中提出的拓扑结构适用于任何规模的数据中心。<ref>{{Cite web |date=2011-11-06 |title=Telecommunications Standards Development |url=http://www.tiaonline.org/standards/ |website=web.archive.org |access-date=2022-10-02 |archive-date=2011-11-06 |archive-url=https://web.archive.org/web/20111106042758/http://www.tiaonline.org/standards/ |dead-url=unfit }}</ref>▼
[[信息安全]]也是一个令人担忧的问题,因此,数据中心必须提供安全的环境,以最大限度地减少安全漏洞的可能性。因此,数据中心必须保持高标准,以确保其托管计算机环境的完整性和功能性。
行业研究公司[[国际数据公司]] (IDC) 估计,数据中心的平均使用年限为9年。<ref name="mspmentor.net"/> 另一家研究公司[[高德纳咨询公司]]表示,使用年限超过7年的数据中心已经过时。<ref>{{cite magazine|url=https://www.forbes.com/2010/03/12/cloud-computing-ibm-technology-cio-network-data-centers.html|title=Sperling, Ed. "Next-Generation Data Centers," Forbes, March 15. 2010|magazine=Forbes.com|access-date=2013-08-30|archive-date=2023-12-10|archive-url=https://web.archive.org/web/20231210145904/https://www.forbes.com/2010/03/12/cloud-computing-ibm-technology-cio-network-data-centers.html|dead-url=no}}</ref> 数据的增长(到2025年将达到 163 [[泽字节]] (ZetaBytes, ZB)<ref>{{Cite web|url=https://www.seagate.com/files/www-content/our-story/trends/files/Seagate-WP-DataAge2025-March-2017.pdf|title=IDC white paper, sponsored by Seagate|access-date=2024-07-28|archive-date=2017-12-08|archive-url=https://web.archive.org/web/20171208010603/https://www.seagate.com/files/www-content/our-story/trends/files/Seagate-WP-DataAge2025-March-2017.pdf|dead-url=no}}</ref>)是推动数据中心现代化需求的一个因素。
关注[[現代化理論|现代化]]并不是什么新鲜事:2007年人们就曾谴责过对过时设备的担忧,<ref>{{cite web|date=December 10, 2007|title=Data centers are aging, unsuited for new technologies|url=https://www.datacenterdynamics.com/news/data-centers-are-aging-unsuited-for-new-technologies-and-businesses-are-slow-to-respond-says-new-ari-survey}}</ref> 2011年,Uptime Institute 也曾担心其中设备的使用年限。<ref group="note">In May 2011, data center research organization Uptime Institute reported that 36 percent of the large companies it surveyed expect to exhaust IT capacity within the next 18 months. {{cite magazine|magazine=[[CIO magazine]]|title=Data Centers Turn to Outsourcing to Meet Capacity Needs|url=http://www.cio.com/article/681897/Data_Centers_Turn_to_Outsourcing_to_Meet_Capacity_Needs|author=James Niccolai|access-date=2011-09-09|archive-date=2011-11-15|archive-url=https://web.archive.org/web/20111115103016/http://www.cio.com/article/681897/Data_Centers_Turn_to_Outsourcing_to_Meet_Capacity_Needs|url-status=dead}}</ref> 到2018年,人们的担忧再次转移,这次是员工的年龄:“数据中心员工的老化速度比设备更快。”<ref>{{cite magazine|magazine=[[Network World]]|title=Data center staff are aging faster than the equipment|url=https://www.networkworld.com/article/966282/data-center-staff-are-aging-faster-than-the-equipment.html|date=August 30, 2018|access-date=December 21, 2018|archive-date=December 7, 2023|archive-url=https://web.archive.org/web/20231207000228/https://www.networkworld.com/article/966282/data-center-staff-are-aging-faster-than-the-equipment.html|url-status=live}}</ref>
=== 满足数据中心标准 ===
▲电信行业协会<ref>{{Cite web |title=Telecommunications Industry Association 942 Certifications & Ratings |url=https://tiaonline.org/products-and-services/tia942certification/tia-942-certifications-ratings/ |website=TIA Online |language=en-US |access-date=2022-10-02 |archive-date=2022-10-02 |archive-url=https://web.archive.org/web/20221002035027/https://tiaonline.org/products-and-services/tia942certification/tia-942-certifications-ratings/ |dead-url=no }}</ref>制定了数据中心电信基础设施标准,规定了数据中心和计算机机房基础设施(包括单租户用户企业数据中心和多租户用户互联网
Telcordia GR-3160(NEBS对电信数据中心设备和空间的要求<ref>{{Cite web |title=GR-3160 - Telecommunications Data Center {{!}} Telcordia |url=https://telecom-info.njdepot.ericsson.net/site-cgi/ido/docs.cgi?ID=SEARCH&DOCUMENT=GR-3160& |website=telecom-info.njdepot.ericsson.net |access-date=2022-10-02 |archive-date=2022-10-07 |archive-url=https://web.archive.org/web/20221007030412/https://telecom-info.njdepot.ericsson.net/site-cgi/ido/docs.cgi?ID=SEARCH&DOCUMENT=GR-3160& |dead-url=no }}</ref>)提供了电信网络内的数据中心空间的指导方针,以及在这些空间中被安装设备的环境要求——这些标准是由Telcordia和行业代表共同制定。其可应用于容纳数据处理或信息技术(IT)设备的数据中心空间。该设备可用于:
第49行 ⟶ 第71行:
不同類別的數據中心會有不同的設計考量,從地點的選擇開始,到外部電力運載需求、物理佈局、冷卻系統 (空調) 的考量、內部供電系統的設計、防火防震自然災害、以及安全防護等等的設計都需要在開始建設數據中心前就考慮決定好。<ref>A ConnectKentucky article mentioning Stone Mountain Data Center Complex {{Cite web |title=Global Data Corp. to Use Old Mine for Ultra-Secure Data Storage Facility |url=http://connectkentucky.org/_documents/connected_fall_FINAL.pdf |accessdate=2007-11-01 |date=2007-11-01 |format=PDF |publisher=ConnectKentucky |dead-url=yes |archive-url=https://web.archive.org/web/20110725192418/http://connectkentucky.org/_documents/connected_fall_FINAL.pdf |archive-date=2011-07-25}}</ref>
==
几十年来,数据中心设计领域一直在向各个方向发展,包括大大小小的新建筑以及对现有设施的创造性再利用,如废弃的零售空间、旧盐矿和战争时期的掩体。
当地建筑规范可能规定了最低天花板高度和其他参数。数据中心设计中的一些考虑因素包括:
[[File:Rack001.jpg|thumb|right|典型的服务器机架,常见于[[主機託管]].]]
* 大小 - 建筑物的一个房间、一个或多个楼层或整个建筑物,
* 容量 - 可容纳多达或超过1,000台服务器<ref>{{Cite web|url=https://www.youtube.com/watch?v=zRwPSFpLX8I| archive-url=https://ghostarchive.org/varchive/youtube/20211104/zRwPSFpLX8I| archive-date=2021-11-04 | url-status=live|title=Google container data center tour (video)| website=[[YouTube]]| date=7 April 2009}}{{cbignore}}</ref>
* 其他考虑因素 - 数据中心的空间、电力、冷却和成本。<ref>{{cite web|url=http://www.networkcomputing.com/data-center/231000669|title=Romonet Offers Predictive Modeling Tool For Data Center Planning|date=June 29, 2011|access-date=February 8, 2012|archive-date=August 23, 2011|archive-url=https://web.archive.org/web/20110823041831/http://www.networkcomputing.com/data-center/231000669|url-status=dead}}</ref>
* 机械工程基础设施 - 供暖、通风和空调 ([[HVAC]]);加湿和除湿设备;增压。<ref name="nxtbook.com">{{cite web |title=BICSI News Magazine - May/June 2010 |url=http://www.nxtbook.com/nxtbooks/bicsi/news_20100506/#/26 |website=www.nxtbook.com |access-date=2024-07-28 |archive-date=2019-04-20 |archive-url=https://web.archive.org/web/20190420132241/http://www.nxtbook.com/nxtbooks/bicsi/news_20100506/#/26 |dead-url=no }}</ref>
* 电气工程基础设施设计 - 电力服务规划;电源的配电、切换和旁路;[[不间断电源]] (UPS) 系统;等等。<ref name="nxtbook.com" /><ref>{{cite web |title=Hedging Your Data Center Power |url=http://www.datacenterjournal.com/design/hedging-your-data-center-power/ |access-date=2024-07-28 |archive-date=2024-05-17 |archive-url=https://web.archive.org/web/20240517235843/https://www.datacenterjournal.com/design/hedging-your-data-center-power/ |dead-url=no }}</ref>
[[File:CRAC Cabinets 2.jpg|thumb|数据中心CRAC冷却]]
=== 物理布局 ===
一个数据中心占用一幢大楼的一个房间,一层或多层,甚至整栋大楼。大部份的设备常常放在具有19英寸的隔层的机架中。这些机架成排放置,形成一个走廊。这允许人们从前面或后面访问隔层。服务器从1U的服务器到独立筒仓的存储设备在尺寸上有很大的不同,存储设备要占掉很多块地砖。一些设备,像大型计算机和存储设备常常像他们的机架那么大,并被放在他们的旁边。非常大的数据中心可以使用集装箱来放置,每个[[集装箱]]可以放置1000或者更多的服务器;当有维修或升级需要的时候,整个集装箱会被替换而不是维修单个的服务器。本地编译的代码可以控制最小的上升限度。
第67行 ⟶ 第104行:
=== 防火 ===
[[File:FM200 Three.jpg|thumb|[[七氟丙烷|FM200]]灭火系统储罐。]]
数据中心采用[[防火]]系统,包括[[被动消防|被动的]]和主动的设计元素,并在业务运行中实施[[防火]]计划。通常会安装[[烟雾探测器]],以便在火灾初期发出预警。
虽然由于[[印制电路板]]的脆弱性,主机房通常不允许使用[[自動灑水系統|湿管的系统]],但仍有一些系统可用于设施的其他部分或冷/热通道空气循环系统中的封闭系统,如<ref>{{Cite web|url=https://www.facilitiesnet.com/datacenters/article/Data-Center-Fire-Suppression-Systems-What-Facility-Managers-Should-Consider--14595|title=Data Center Fire Suppression Systems: What Facility Managers Should Consider|website=Facilitiesnet|access-date=2024-07-28|archive-date=2024-05-22|archive-url=https://web.archive.org/web/20240522091545/https://www.facilitiesnet.com/datacenters/article/Data-Center-Fire-Suppression-Systems-What-Facility-Managers-Should-Consider--14595|dead-url=no}}</ref>:
* [[自動灑水系統]]
* [[薄霧|喷雾]],利用高压产生极小的水滴,由于水滴的性质,可用于敏感的房间。
不过,也有其他[[灭火]]方法,特别是在[[服务器农场|敏感地区]],通常使用[[气体灭火]],其中[[三氟溴甲烷]]气体(哈龙1301, Halon 1301)最受欢迎,直到发现其生产和使用的负面影响。[https://www.epa.gov/ozone-layer-protection/halons-program] {{Wayback|url=https://www.epa.gov/ozone-layer-protection/halons-program |date=20240305040334 }}
=== 保安 ===
物理访问通常受到限制。分层安全通常从围栏、系缆桩和门禁系统开始<ref>{{cite web|author=Sarah D. Scalet|url=http://www.csoonline.com/article/220665|title=19 Ways to Build Physical Security Into a Data Center|publisher=Csoonline.com|date=2005-11-01|access-date=2013-08-30|archive-date=2008-04-21|archive-url=https://web.archive.org/web/20080421020352/http://www.csoonline.com/article/220665|url-status=dead}}</ref>。 如果数据中心规模较大或包含敏感信息,则几乎总会有[[攝影機]]监控和长期[[保全|保安員]]。指纹识别门禁系统已开始普及。
一些数据保护法规要求对访问进行记录;一些组织将此与访问控制系统紧密联系在一起。可在主入口、内部房间入口和设备机柜处输入多个日志。机柜的门禁控制可与智能配电装置集成,从而通过同一设备将锁联网。<ref>{{Citation|title=Systems and methods for controlling an electronic lock for a remote device|date=2016-08-01|url=https://patents.google.com/patent/US9865109B2/en|access-date=2018-04-25|archive-date=2023-03-06|archive-url=https://web.archive.org/web/20230306024610/https://patents.google.com/patent/US9865109B2/en|dead-url=no}}</ref>
== 流動數據中心 ==
第79行 ⟶ 第127行:
生產流動數據中心的廠商和產品包括:
* [[Google]]的Google Modular Data Center<ref>{{YouTube|id=zRwPSFpLX8I|title=Google container data center tour}}</ref><ref>{{Cite web |title=IBM's Project Big Green Takes Second Step |url=http://www.crn.com/hardware/208403225 |accessdate=2008-05-11 |date=June 11, 2008 |last=Kraemer |first=Brian |publisher=ChannelWeb |archiveurl=https://web.archive.org/web/20080611114732/http://www.crn.com/hardware/208403225 |archivedate=2008年6月11日 |deadurl=yes}}</ref><ref>{{Cite web |title=Modular/Container Data Centers Procurement Guide: Optimizing for Energy Efficiency and Quick Deployment |url=http://hightech.lbl.gov/documents/data_centers/modular-dc-procurement-guide.pdf |accessdate=2013-08-30 |format=PDF |archiveurl=https://web.archive.org/web/20130531191212/http://hightech.lbl.gov/documents/data_centers/modular-dc-procurement-guide.pdf |archivedate=2013-05-31 |deadurl=yes}}</ref>、
* [[昇陽電腦]](已被甲骨文收购)的Sun Modular Datacenter <ref>{{Cite web |title=Sun Modular Datacenter S20 and D20 Documentation |url=http://docs.oracle.com/cd/E19115-01/mod.dc.s20/index.html |accessdate=2012-02-12 |dead-url=no |archive-date=2012-02-12 |archive-url=https://web.archive.org/web/20120212000533/http://docs.oracle.com/cd/E19115-01/mod.dc.s20/index.html}}</ref>、
* [[IBM]]的Portable Modular Data Center<ref>{{Cite web |title=IBM Portable Modular Data Center |url=http://www-935.ibm.com/services/us/en/it-services/it-facilities-assessment-design-and-construction-services-portable-modular-data-center.html |accessdate=2011-09-29 |dead-url=no |archive-date=2011-10-08 |archive-url=https://web.archive.org/web/20111008032516/http://www-935.ibm.com/services/us/en/it-services/it-facilities-assessment-design-and-construction-services-portable-modular-data-center.html}}</ref>
* [[思科系统]]的Containerized Data Center
第94行 ⟶ 第142行:
==參见==
{{columns-list|colwidth=15em|
*[[服务器农场]]▼
* [[電腦硬體冷卻]]
* [[灾难恢复]]
* [[电路]]
* [[互联网交换中心]]
* [[互联网托管服务]]
* [[网络运维中心]]
* [[开放计算项目]] - [[Meta Platforms|Facebook]]
* [[Peering]]
▲* [[服务器农场]]
* [[服务器机房]]
* [[设施管理]]
* [[电信网络]]
* [[犹他数据中心]]
* [[網頁寄存服務]]
}}
== 參考文獻 ==
{{Reflist|30em}}
== 注释 ==
{{Reflist|group=note}}
{{雲端運算}}
{{-}}
|