外文翻譯---超越臺(tái)式機(jī)一個(gè)關(guān)于云計(jì)算的介紹

1、<p><b>　　附錄A 譯文</b></p><p>　　超越臺(tái)式機(jī)：一個(gè)關(guān)于云計(jì)算的介紹 </p><p>　　在這個(gè)世界上幾乎每一天都能看到新技術(shù)趨勢的崛起和衰退，一種新的趨勢帶來了更長久的生命力。這種趨勢被稱為云計(jì)算，它將改變你使用電腦和互聯(lián)網(wǎng)的方式。</p><p>　　云計(jì)算預(yù)示著我們在如何存儲(chǔ)信息和運(yùn)行應(yīng)用程序問題上

2、的重大變革。而不是在單獨(dú)的計(jì)算機(jī)上運(yùn)行程序的散的數(shù)據(jù)，一切都托管在“云”中，通過英特網(wǎng)去訪問“云”中的其他計(jì)算機(jī)和服務(wù)器。云計(jì)算讓你能夠在世界上的任何地方訪問你所有的應(yīng)用程序和文檔，擺脫了臺(tái)式機(jī)的束縛，讓你更容易在不同的地方和同事進(jìn)行協(xié)作。</p><p>　　第一部分 認(rèn)識(shí)云計(jì)算</p><p>　　云計(jì)算的出現(xiàn)就相當(dāng)于一個(gè)世紀(jì)前電力革命的發(fā)生。在電力公司出現(xiàn)之前，每一個(gè)農(nóng)場和企業(yè)都通過

3、自己獨(dú)立的發(fā)電機(jī)發(fā)電。在電網(wǎng)產(chǎn)生以后，農(nóng)場和企業(yè)關(guān)閉他們自己的的發(fā)電機(jī)，并且從電力公司以一個(gè)比他們用自己發(fā)電機(jī)發(fā)電低得多的價(jià)格（并且更具有可靠性）買電。</p><p>　　為了尋找相同類型革命的發(fā)生，使得云計(jì)算成為定局。由于我們所期望的教育普及，每時(shí)每刻的可靠性和無處不在的云計(jì)算協(xié)作承諾使得以臺(tái)式機(jī)為計(jì)算中心的觀念半途而廢。云計(jì)算將是未來發(fā)展的方向。</p><p>　　云計(jì)算：它是什么

4、和它的不同</p><p>　　通過傳統(tǒng)的桌面計(jì)算，你可以在你的每一臺(tái)電腦上運(yùn)行程序的副本，你所創(chuàng)建的文件都保存在你創(chuàng)建文件的計(jì)算機(jī)上。雖然它們能夠被這個(gè)網(wǎng)絡(luò)上的其他計(jì)算機(jī)訪問，但是它們不能被外網(wǎng)的計(jì)算機(jī)訪問。</p><p>　　整個(gè)的場景都是以個(gè)人計(jì)算機(jī)為中心。</p><p>　　通過云計(jì)算，你使用的軟件程序不需要運(yùn)行在你的個(gè)人電腦上，但是卻可以通過訪問英特網(wǎng)

5、上的服務(wù)器使用這個(gè)軟件程序。假如你的電腦死機(jī)了，但是這個(gè)軟件程序仍然可以供其他人使用。這種模式仍然適用于你創(chuàng)建的文件，它們被儲(chǔ)存在一個(gè)可以通過互聯(lián)網(wǎng)訪問的集合中。經(jīng)過許可的任何人不僅可以訪問文件，還可以實(shí)時(shí)地編輯和處理這些文件。不同于傳統(tǒng)的計(jì)算，云計(jì)算模式不以個(gè)人計(jì)算機(jī)為中心，而是以文檔為中心，個(gè)人計(jì)算機(jī)使用簡單的文件訪問并不重要。</p><p>　　但是這只是一個(gè)簡化。讓我們來看看更詳細(xì)的云計(jì)算是什么樣子的？

6、它和我們了解云計(jì)算的不同同樣重要。</p><p><b>　　云計(jì)算的不同</b></p><p>　　首先，云計(jì)算不是網(wǎng)絡(luò)計(jì)算。通過網(wǎng)絡(luò)計(jì)算，應(yīng)用程序和文檔被托管在一個(gè)公司的服務(wù)器中并且通過公司的網(wǎng)絡(luò)可以訪問那些應(yīng)用程序和文檔。云計(jì)算要比網(wǎng)絡(luò)計(jì)算大很多，它包含了許多個(gè)公司，許多個(gè)服務(wù)器和許多個(gè)網(wǎng)絡(luò)。此外，與網(wǎng)絡(luò)計(jì)算不同，云服務(wù)和儲(chǔ)存通過網(wǎng)絡(luò)在世界的任何地方都可以訪

7、問，但是通過網(wǎng)絡(luò)計(jì)算，只能在自己公司的網(wǎng)絡(luò)訪問到。</p><p>　　云計(jì)算也不是傳統(tǒng)的像一個(gè)公司將計(jì)算服務(wù)外包（分包）給一個(gè)外部公司。雖然外包公司可以托管一個(gè)公司的數(shù)據(jù)或者應(yīng)用程序，但是這些程序和計(jì)劃只有這個(gè)公司的員工通過公司的網(wǎng)絡(luò)才能訪問到，而不是通過整個(gè)英特網(wǎng)。</p><p>　　因此，盡管表面上很相似，但是網(wǎng)絡(luò)計(jì)算和外包不是云計(jì)算。</p><p>&l

8、t;b>　　什么是云計(jì)算</b></p><p>　　云計(jì)算定義的關(guān)鍵是“云”本身。對于我們而言，云是一大群互聯(lián)的計(jì)算機(jī)。這些計(jì)算機(jī)可以是個(gè)人電腦或者網(wǎng)絡(luò)服務(wù)器，它們可以是公共或者私人的。</p><p>　　例如，谷歌的云是由小的個(gè)人電腦和更大的服務(wù)器組成的。谷歌云是私人的（即，谷歌擁有它），是公開訪問的（通過谷歌用戶）。</p><p>　　這

9、種計(jì)算機(jī)云繼承并且超越了單一的公司或者企業(yè)?？缙髽I(yè)的跨平臺(tái)的廣大用戶群可以訪問云服務(wù)所提供的應(yīng)用程序和數(shù)據(jù)。通過互聯(lián)網(wǎng)訪問，任何被授權(quán)的用戶可以訪問連接在英特網(wǎng)上的任何計(jì)算機(jī)上的這些文檔和應(yīng)用程序。并且，對于用戶而言，云技術(shù)和云背后的基礎(chǔ)設(shè)施是不可見的。</p><p>　　云技術(shù)是否基于HTTP,HTML協(xié)議，XML，JavaScript或者其他特定的技術(shù)是不明顯的。（在大多數(shù)情況下是無所謂的）。</p&

10、gt;<p>　　云計(jì)算是用戶為中心的。一旦你作為一個(gè)用戶連接到云，無論儲(chǔ)存在哪里的文件，信息，圖像，應(yīng)用程序，無論什么都將變成你的。此外，你的數(shù)據(jù)也可以和他人分享。事實(shí)上，任何訪問你的數(shù)據(jù)的設(shè)備在云中都將變成你的。</p><p>　　云計(jì)算是以任務(wù)為中心的。而不是以應(yīng)用程序和它能為你做什么為重點(diǎn)，重點(diǎn)是你需要做什么和應(yīng)用程序怎樣為你完成這個(gè)需求。傳統(tǒng)的應(yīng)用程序----文字處理，電子表格，電子郵件

11、等，都將變得沒有它們所產(chǎn)生的文件重要。</p><p>　　第二部分 認(rèn)識(shí)云計(jì)算</p><p>　　云計(jì)算是強(qiáng)大的。連接在一起的數(shù)百臺(tái)或者數(shù)千臺(tái)計(jì)算機(jī)共同創(chuàng)造的計(jì)算能力是一臺(tái)單獨(dú)的臺(tái)式計(jì)算機(jī)無法比擬的。</p><p>　　云計(jì)算是可訪問的。因?yàn)閿?shù)據(jù)被存儲(chǔ)在云中，用戶可以迅速地從多個(gè)庫中檢索出更多的信息，你不僅限于一個(gè)單獨(dú)的數(shù)據(jù)源，如同你和你的臺(tái)式個(gè)人電腦一樣。&

12、lt;/p><p>　　云計(jì)算是智能的。由于云中電腦儲(chǔ)存著各種數(shù)據(jù)，數(shù)據(jù)挖掘和分析對于以智能的方式訪問這些信息是必要的。</p><p>　　云計(jì)算是可編程的。許多伴隨云計(jì)算的必要任務(wù)必須要實(shí)現(xiàn)自動(dòng)化。例如，要保護(hù)存儲(chǔ)在云中某臺(tái)計(jì)算機(jī)上的集成數(shù)據(jù)和信息需要在云中的另外一臺(tái)計(jì)算機(jī)上完整的復(fù)制。如果云中的某臺(tái)準(zhǔn)備被用于復(fù)制這些數(shù)據(jù)和信息的計(jì)算機(jī)脫機(jī)，那么云的編程將自動(dòng)的將該計(jì)算機(jī)的數(shù)據(jù)復(fù)制到云中新

13、的計(jì)算機(jī)上。</p><p>　　所有的這些都是在我們背后定義的，那么什么才是現(xiàn)實(shí)世界中的云計(jì)算？</p><p>　　當(dāng)你學(xué)習(xí)完整本書，網(wǎng)站托管，互聯(lián)網(wǎng)訪問，團(tuán)體協(xié)作的應(yīng)用程序一般是可用的。伴隨著更多的方式，也許今天最好最流行的云計(jì)算應(yīng)用例子就是谷歌大家庭中的應(yīng)用程序---谷歌文檔和電子表格，谷歌日歷，谷歌郵箱，圖片管理等等。這些應(yīng)用都被托管在谷歌的服務(wù)器中，是任何連接在互聯(lián)網(wǎng)上的用戶都

14、可以訪問的，可用于在世界不同地域的團(tuán)體協(xié)作。</p><p>　　總之，云計(jì)算使計(jì)算機(jī)轉(zhuǎn)移到用戶，從應(yīng)用到任務(wù)，從孤立的數(shù)據(jù)到任何地方都可以訪問到的數(shù)據(jù)，并且與任何人共享。用戶不再承擔(dān)管理數(shù)據(jù)的任務(wù)，他甚至不需要記得數(shù)據(jù)在哪里。所有的事實(shí)是數(shù)據(jù)在云中，對于用戶和被授權(quán)的用戶是可以迅速訪問的。</p><p>　　從協(xié)作到云：云計(jì)算的簡史</p><p>　　云計(jì)算的

15、前身是客戶機(jī)/服務(wù)器模式的計(jì)算和點(diǎn)對點(diǎn)的分布式計(jì)算。問題是如何集中存儲(chǔ)協(xié)作和如何使多臺(tái)計(jì)算機(jī)協(xié)同工作以增加計(jì)算能力。</p><p>　　客戶機(jī)/服務(wù)器模式計(jì)算：集中應(yīng)用和存儲(chǔ)</p><p>　　很久以前的計(jì)算（1980年左右）,一切的操作都是客戶機(jī)/服務(wù)器模式。所有的應(yīng)用程序，所有的數(shù)據(jù)，都被放在大型計(jì)算機(jī)上，這種大型計(jì)算機(jī)被稱為服務(wù)器。如果用戶想訪問特定的數(shù)據(jù)或者運(yùn)行程序，就必要連接

16、到大型計(jì)算機(jī)，獲得適當(dāng)?shù)脑L問，然后從大型計(jì)算機(jī)上得到程序或者數(shù)據(jù)以完成相應(yīng)的操作。</p><p>　　用戶通過計(jì)算機(jī)終端連接到服務(wù)器，有時(shí)也被稱為工作站或者客戶端。這臺(tái)計(jì)算機(jī)有時(shí)被稱為啞終端，因?yàn)樗鼪]有足夠（如果有的話）的內(nèi)存，存儲(chǔ)空間，處理能力。這僅僅是連接到一個(gè)用戶，使他能夠使用計(jì)算機(jī)主機(jī)設(shè)備。</p><p>　　當(dāng)用戶獲準(zhǔn)訪問大型主機(jī)時(shí)，信息技術(shù)(IT)的工作人員沒有偶然訪問時(shí)不

17、在的習(xí)慣。甚至在一個(gè)大型主機(jī)上，處理能力也是有限的---IT工作人員有監(jiān)護(hù)這種處理能力的權(quán)利。這種訪問不是立竿見影的，也不可能有兩個(gè)用戶同時(shí)訪問相同的數(shù)據(jù)。</p><p>　　除此之外，用戶幾乎不得不采取IT工作人員給他們的一切措施（沒有任何變化）。想要使用自定義報(bào)告去顯示正常的信息子集？但是你不能這么做，雖然IT工作人員可以，但是這是他們從現(xiàn)在開始到數(shù)周后的日程安排。</p><p>

18、　　事實(shí)是，當(dāng)許多人共用一臺(tái)計(jì)算機(jī)，即使這是一臺(tái)大型主機(jī)，你也不得不排隊(duì)等候。需要返回一個(gè)財(cái)務(wù)報(bào)告？沒問題，如果你不介意一直等到今天下午或者明天早上。在客戶機(jī)/服務(wù)器模式環(huán)境中總是沒有直接訪問的，很少滿足直接訪問的情況。</p><p>　　因此客戶機(jī)/服務(wù)器模式，雖然提供類似的集中存儲(chǔ)，但是不同于以用戶為中心的云計(jì)算，在客戶機(jī)/服務(wù)器模式計(jì)算中，所有的控制權(quán)在大型主機(jī)和控制該臺(tái)計(jì)算機(jī)上的人手中。這不是一個(gè)有利于

19、用戶的環(huán)境。</p><p>　　點(diǎn)對點(diǎn)計(jì)算：資源共享</p><p>　　你可以想象，當(dāng)問客戶機(jī)/服務(wù)器模式的系統(tǒng)是一種“趕快和等待”的經(jīng)歷。該系統(tǒng)的服務(wù)器部分還創(chuàng)建了龐大的阻礙。計(jì)算機(jī)之間的所有通信必須先通過服務(wù)器，可是這種方式可能效率太低。</p><p>　　最明顯的是由于一臺(tái)計(jì)算機(jī)連接到另一臺(tái)計(jì)算機(jī)需要通過服務(wù)器促使了點(diǎn)對點(diǎn)（P2P）計(jì)算的發(fā)展。P2P計(jì)算

20、定義了網(wǎng)絡(luò)中每一臺(tái)計(jì)算機(jī)具有同等責(zé)任和能力的網(wǎng)絡(luò)架構(gòu)。這種相對于傳統(tǒng)的客戶機(jī)/服務(wù)器模式的網(wǎng)絡(luò)體系架構(gòu)，在網(wǎng)絡(luò)中有一臺(tái)或者多臺(tái)計(jì)算機(jī)是專門為其他計(jì)算提供服務(wù)的。（這種關(guān)系有時(shí)有主/從關(guān)系的特點(diǎn)，中央服務(wù)器是主，客戶機(jī)是從）。</p><p>　　P2P是一個(gè)均衡的概念。在P2P環(huán)境中，每臺(tái)計(jì)算機(jī)是一個(gè)客戶機(jī)和服務(wù)器，沒有主和從。通過承認(rèn)所有計(jì)算機(jī)在網(wǎng)絡(luò)中的相同地位，P2P使得網(wǎng)絡(luò)中的計(jì)算機(jī)可以直接交換資源和提供相

21、互服務(wù)。沒有中央處理器的需求，因?yàn)槊恳慌_(tái)計(jì)算機(jī)都能在要求這樣做時(shí)提供這樣的功能。</p><p>　　點(diǎn)對點(diǎn)也是分散的概念。控制是分散的，所有計(jì)算機(jī)的功能是平等的。內(nèi)容在不同計(jì)算機(jī)之間也是分散的。沒有集中服務(wù)器被分配到主機(jī)的可用資源和服務(wù)。</p><p>　　也許最引人注目的P2P計(jì)算實(shí)施是英特網(wǎng)。今天的許多用戶都忘記（也許不知道），英特網(wǎng)最初的設(shè)想是在其最初ARPA網(wǎng)的幌子下，作為一個(gè)

22、能夠分享在美國的計(jì)算資源的點(diǎn)對點(diǎn)系統(tǒng)。ARPA網(wǎng)的各種網(wǎng)站（其中的許多不屬于他們），連接在一起的計(jì)算機(jī)不是作為服務(wù)器和客戶機(jī)，而是平等的。</p><p>　　初期的P2P英特網(wǎng)最好例子是Usenet網(wǎng)絡(luò)。它創(chuàng)建于1979年，是一個(gè)計(jì)算機(jī)網(wǎng)絡(luò)（通過英特網(wǎng)訪問）,每一臺(tái)計(jì)算機(jī)提供該網(wǎng)絡(luò)的全部內(nèi)容。消息在平等的計(jì)算機(jī)之間傳播。用戶連接到Usenet的每一個(gè)服務(wù)器獲得所有（或者幾乎所有）的消息以張貼到每個(gè)單獨(dú)的服務(wù)器。

23、雖然用戶連接到Usenet服務(wù)器是傳統(tǒng)客戶機(jī)/服務(wù)器模式的性質(zhì)，但是這種Usenet服務(wù)器之間的關(guān)系肯定了P2P，也預(yù)示著今天的云計(jì)算。</p><p>　　不過，并不是所有的英特網(wǎng)的一部分就是P2P性質(zhì)。隨著萬維網(wǎng)中P2P的發(fā)展，退后了客戶機(jī)/服務(wù)器模式。在網(wǎng)絡(luò)上，每一個(gè)網(wǎng)站是由一個(gè)計(jì)算機(jī)組提供服務(wù)，網(wǎng)站的訪問者使用客戶端軟件（網(wǎng)絡(luò)瀏覽器）來訪問它。幾乎所有的內(nèi)容都被集中，所有的控制都被集中，在這個(gè)過程中，客戶

24、端沒有自主權(quán)或者控制權(quán)在。</p><p>　　分布式計(jì)算：提供更強(qiáng)的計(jì)算能力</p><p>　　P2P模型中最重要的子集之一是分布式計(jì)算，通過個(gè)人計(jì)算機(jī)在網(wǎng)絡(luò)或者在英特網(wǎng)上的閑置來提供大型的，處理器密集型項(xiàng)目的計(jì)算能力。所有關(guān)于多臺(tái)計(jì)算機(jī)間的周期的共享是一個(gè)簡單的概念。</p><p>　　個(gè)人計(jì)算機(jī)，每天運(yùn)行滿24小時(shí)，每周7天，是巨大的計(jì)算能力。大多數(shù)人不每

25、時(shí)每刻使用他們電腦，卻除去這樣一個(gè)電腦的閑置資源。分布式計(jì)算使用這些資源。</p><p>　　當(dāng)一臺(tái)計(jì)算機(jī)中加入到一個(gè)分布式的計(jì)算項(xiàng)目，軟件被安裝在這臺(tái)計(jì)算機(jī)上在用戶未使用的這段時(shí)間去運(yùn)行各種處理活動(dòng)。這個(gè)業(yè)余處理的結(jié)果被定期上傳到分布式計(jì)算網(wǎng)絡(luò)上，并與項(xiàng)目中其他計(jì)算機(jī)類似的結(jié)果相結(jié)合。結(jié)果是，如果有足夠的計(jì)算機(jī)被涉及，就模擬了更大的大型主機(jī)和超級計(jì)算機(jī)的處理能力，這對于某些大型和復(fù)雜的計(jì)算項(xiàng)目的必要的。<

26、;/p><p>　　例如，基因研究需要大量的計(jì)算能力。對于傳統(tǒng)手段，它可能需要數(shù)年的時(shí)間來解決基本的數(shù)學(xué)問題。通過連接在一起的數(shù)千（或者百萬）臺(tái)個(gè)人計(jì)算機(jī)，更多的能力適用于這個(gè)問題，得出結(jié)論就更快。</p><p>　　分布式計(jì)算的歷史可以追溯到1973年，多臺(tái)計(jì)算機(jī)聯(lián)網(wǎng)的施樂公司帕洛阿爾托研究中心是通過網(wǎng)絡(luò)尋找閑置資源來開發(fā)殺毒軟件的。一個(gè)分布式計(jì)算更實(shí)際的應(yīng)用出現(xiàn)在1988年，當(dāng)DEC（數(shù)

27、字設(shè)備公司）系統(tǒng)研究中心的研究人員開發(fā)軟件，在實(shí)驗(yàn)室中分布式的工作應(yīng)用在各個(gè)工作組之間以實(shí)現(xiàn)大數(shù)字。到1990年，大約100個(gè)用戶組，利用這個(gè)軟件實(shí)現(xiàn)100位數(shù)字，到1995年，同樣的效果被擴(kuò)展到網(wǎng)絡(luò)來實(shí)現(xiàn)130位數(shù)字。</p><p>　　沒過多久分布式計(jì)算進(jìn)入了互聯(lián)網(wǎng)。第一個(gè)主要的基于英特網(wǎng)的的分布式計(jì)算項(xiàng)目是distribute.net，在1997年推出，雇傭了成千上萬的個(gè)人計(jì)算機(jī)來破譯加密代碼。即使是較大

28、的SETI@home，1999年5月，推出了使用連接在一起的數(shù)百萬臺(tái)的個(gè)人計(jì)算機(jī)搜索外層空間智慧生命。許多分布式計(jì)算項(xiàng)目是在大型企業(yè)內(nèi)進(jìn)行，采用傳統(tǒng)的網(wǎng)絡(luò)連接，形成分布式計(jì)算網(wǎng)絡(luò)。另外，更大的項(xiàng)目通常使用互聯(lián)網(wǎng)用戶的計(jì)算機(jī)，通常發(fā)生脫機(jī)，然后每天通過傳統(tǒng)的消費(fèi)英特網(wǎng)上傳資料。</p><p><b>　　了解云架構(gòu)</b></p><p>　　云計(jì)算的關(guān)鍵是“云”，一

29、些大規(guī)模網(wǎng)絡(luò)服務(wù)器或者個(gè)人計(jì)算機(jī)連接在龐大的網(wǎng)絡(luò)中。這些計(jì)算機(jī)并行運(yùn)行，結(jié)合各自的資源，產(chǎn)生超級計(jì)算般的能力。</p><p>　　究竟什么是“云”？簡而言之，云是計(jì)算機(jī)和服務(wù)器都通過英特網(wǎng)公開訪問的集合。這種硬件通常是復(fù)合的和由有一個(gè)或者多個(gè)數(shù)據(jù)中心為穩(wěn)固基礎(chǔ)的第三方運(yùn)作。這些機(jī)器可以運(yùn)行任何操作系統(tǒng)，這些機(jī)器的處理能力不是它們作為單獨(dú)臺(tái)式機(jī)所能比擬的。</p><p>　　如圖1.1所

30、示，個(gè)人用戶使用他們自己的計(jì)算機(jī)或者便攜設(shè)備通過英特網(wǎng)連接到云。對于這些個(gè)人用戶，云被看作是一個(gè)單一的應(yīng)用程序，設(shè)備或者文件。云中的硬件（操作系統(tǒng)管理硬件連接）是無形的。</p><p>　　這看似簡單的云架構(gòu)，它確實(shí)需要一些智能化的管理去將這些計(jì)算機(jī)連接在一起，并分配任務(wù)處理以滿足眾多的用戶。正如你可以看到在圖1.2中，這一切都開始于個(gè)別用戶看到前端界面，選擇一個(gè)任務(wù)或者服務(wù)（無論是啟動(dòng)應(yīng)用程序或者打開一個(gè)文檔

31、），用戶的這些請求被傳遞到管理系統(tǒng)，然后找出正確的資源，然后調(diào)用系統(tǒng)的恰當(dāng)資源去供應(yīng)服務(wù)。這些服務(wù)在云中開拓中必要的資源，發(fā)布恰當(dāng)?shù)木W(wǎng)頁應(yīng)用程序，或者創(chuàng)建文件和打開請求的文件，網(wǎng)頁應(yīng)用程序啟動(dòng)后，系統(tǒng)的檢測和計(jì)量功能跟蹤云的使用，使資源分配和使用于正確的用戶。</p><p>　　正如你所看到的，云計(jì)算概念的觀念是許多管理任務(wù)的自動(dòng)化。如果一個(gè)系統(tǒng)需要人管理資源的分配過程那么這個(gè)系統(tǒng)不是云。在這種情況下，該系統(tǒng)是

32、僅僅以老式數(shù)據(jù)中心為基礎(chǔ)的客戶機(jī)/服務(wù)器模式計(jì)算的21世紀(jì)的版本。對于系統(tǒng)實(shí)現(xiàn)云的狀態(tài)，必須更換人工管理實(shí)現(xiàn)自動(dòng)化程序。</p><p><b>　　附錄B 外文原文</b></p><p>　　Beyond the Desktop: An Introduction to Cloud Computing</p><p>　　In a worl

33、d that sees new technological trends bloom and fade on almost a daily basis, one new trend promises more longevity. This trend is called cloud computing, and it will change the way you use your computer and the Internet.

34、</p><p>　　Cloud computing portends a major change in how we store information and run applications. Instead of running program sand data on an individual desktop computer, everything is hosted in the “cloud”

35、—a nebulous assemblage of computers and servers accessed via the Internet. Cloud computing lets you access all your applications and documents from anywhere in the world, freeing you from the confines of the desktop and

36、making it easier for group members in different locations to collaborate.</p><p>　　PART 1 Understanding Cloud Computing </p><p>　　The emergence of cloud computing is the computing equivalent of

37、the electricity revolution of a century ago. Before the advent of electrical utilities, every farm and business produced its own electricity from freestanding generators. After the electrical grid was created, farms and

38、businesses shut down their generators and bought electricity from the utilities, at a much lower price (and with much greater reliability) than they could produce on their own.</p><p>　　Look for the same typ

39、e of revolution to occur as cloud computing takes hold. The desktop-centric notion of computing that we hold today is bound to fall by the wayside as we come to expect the universal access, 24/7 reliability, and ubiquito

40、us collaboration promised by cloud computing. It is the way of the future.</p><p>　　Cloud Computing: What It Is—and What It Isn’t</p><p>　　With traditional desktop computing, you run copies of s

41、oftware programs on each computer you own. The documents you create are stored on the computer on which they were created. Although documents can be accessed from other computers on the network, they can’t be accessed by

42、 computers outside the network.</p><p>　　The whole scene is PC-centric.</p><p>　　With cloud computing, the software programs you use aren’t run from your personal computer, but are rather stored

43、 on servers accessed via the Internet. If your computer crashes, the software is still available for others to use. Same goes for the documents you create; they’re stored on a collection of servers accessed via the Inter

44、net. Anyone with permission can not only access the documents, but can also edit and collaborate on those documents in real time. Unlike traditional computing, this cl</p><p>　　isn’t important.</p>&l

45、t;p>　　But that’s a simplification. Let’s look in more detail at what cloud computing</p><p>　　is—and, just as important, what it isn’t.</p><p>　　What Cloud Computing Isn’t</p><p>

46、;　　First, cloud computing isn’t network computing. With network computing, applications/documents are hosted on a single company’s server and accessed over the company’s network. Cloud computing is a lot bigger than that

47、. It encompasses multiple companies, multiple servers, and multiple networks. Plus, unlike network computing, cloud services and storage are accessible from anywhere in the world over an Internet connection; with network

48、 computing, access is over the company’s network only.</p><p>　　Cloud computing also isn’t traditional outsourcing, where a company farms out (subcontracts) its computing services to an outside firm. While a

49、n outsourcing firm might host a company’s data or applications, those documents and programs are only accessible to the company’s employees via the company’s network, not to the entire world via the Internet.</p>

50、<p>　　So, despite superficial similarities, networking computing and outsourcing are not cloud computing.</p><p>　　What Cloud Computing Is</p><p>　　Key to the definition of cloud computing

51、is the “cloud” itself. For our purposes, the cloud is a large group of interconnected computers. These computers can be personal computers or network servers; they can be public or private.</p><p>　　For exam

52、ple, Google hosts a cloud that consists of both smallish PCs and larger servers. Google’s cloud is a private one (that is, Google owns it) that is publicly accessible (by Google’s users).</p><p>　　This cloud

53、 of computers extends beyond a single company or enterprise. The applications and data served by the cloud are available to broad group of users, cross-enterprise and cross-platform. Access is via the Internet. Any autho

54、rized user can access these docs and apps from any computer over any Internet connection. And, to the user, the technology and infrastructure behind the cloud is invisible.</p><p>　　It isn’t apparent (and, i

55、n most cases doesn’t matter) whether cloud services are based on HTTP, HTML, XML, JavaScript, or other specific technologies.</p><p>　　_ Cloud computing is user-centric. Once you as a user are connected to t

56、he cloud, whatever is stored there—documents, messages, images, applications, whatever—becomes yours. In addition, not only is the data yours, but you can also share it with others. In effect, any device that accesses yo

57、ur data in the cloud also becomes yours.</p><p>　　_ Cloud computing is task-centric. Instead of focusing on the application and what it can do, the focus is on what you need done and how the application can

58、do it for you. Traditional applications—word processing, spreadsheets, email, and so on—are becoming less important than the documents they create.</p><p>　　PART 2 Understanding Cloud Computing</p>&l

59、t;p>　　_ Cloud computing is powerful. Connecting hundreds or thousands of computers together in a cloud creates a wealth of computing power impossible with a single desktop PC.</p><p>　　_ Cloud computing i

60、s accessible. Because data is stored in the cloud, users can instantly retrieve more information from multiple repositories. You’re not limited to a single source of data, as you are with a desktop PC.</p><p&g

61、t;　　_ Cloud computing is intelligent. With all the various data stored on the computers in a cloud, data mining and analysis are necessary to access that information in an intelligent manner.</p><p>　　_ Clou

62、d computing is programmable. Many of the tasks necessary with cloud computing must be automated. For example, to protect the integrity of the data, information stored on a single computer in the cloud must be replicated

63、on other computers in the cloud. If that one computer goes offline, the cloud’s programming automatically redistributes</p><p>　　that computer’s data to a new computer in the cloud.</p><p>　　All

64、 these definitions behind us, what constitutes cloud computing in the real world?</p><p>　　As you’ll learn throughout this book, a raft of web-hosted, Internet-accessible,</p><p>　　Group-collabo

65、rative applications are currently available, with many more on the way. Perhaps the best and most popular examples of cloud computing applications today are the Google family of applications—Google Docs & Spreadsheet

66、s, Google Calendar, Gmail, Picasa, and the like. All of these applications are hosted on Google’s servers, are accessible to any user with an Internet connection, and can be used for group collaboration from anywhere in

67、the world.</p><p>　　In short, cloud computing enables a shift from the computer to the user, from applications to tasks, and from isolated data to data that can be accessed from anywhere and shared with anyo

68、ne. The user no longer has to take on the task of data management; he doesn’t even have to remember where the data is. All that matters is that the data is in the cloud, and thus immediately available to that user and to

69、 other authorized users.</p><p>　　From Collaboration to the Cloud: A Short History of Cloud Computing</p><p>　　Cloud computing has as its antecedents both client/server computing and peer-to-pee

70、r distributed computing. It’s all a matter of how centralized storage facilitates collaboration and how multiple computers work together to increase computing power.</p><p>　　Client/Server Computing: Central

71、ized Applications and Storage</p><p>　　In the antediluvian days of computing (pre-1980 or so), everything operated on the client/server model. All the software applications, all the data, and all the control

72、 resided on huge mainframe computers, otherwise known as servers. If a user wanted to access specific data or run a program, he had to connect to the mainframe, gain appropriate access, and then do his business while ess

73、entially “renting” the program or data from the server.</p><p>　　Users connected to the server via a computer terminal, sometimes called a workstation or client. This computer was sometimes called a dumb ter

74、minal because it didn’t have a lot (if any!) memory, storage space, or processing power. It was merely a device that connected the user to and enabled him to use the mainframe computer.</p><p>　　Users access

75、ed the mainframe only when granted permission, and the information technology (IT) staff weren’t in the habit of handing out access casually. Even on a mainframe computer, processing power is limited—and the IT staff wer

76、e the guardians of that power. Access was not immediate, nor could two users access the same data at the same time.</p><p>　　Beyond that, users pretty much had to take whatever the IT staff gave them—with no

77、 variations. Want to customize a report to show only a subset of the normal information? Can’t do it. Want to create a new report to look at some new data? You can’t do it, although the IT staff can—but on their schedule

78、, which might be weeks from now.</p><p>　　The fact is, when multiple people are sharing a single computer, even if that computer is a huge mainframe, you have to wait your turn. Need to rerun a financial rep

79、ort? No problem—if you don’t mind waiting until this afternoon, or tomorrow morning. There isn’t always immediate access in a client/server environment, and seldom is there immediate gratification.</p><p>　　

80、So the client/server model, while providing similar centralized storage, differed</p><p>　　from cloud computing in that it did not have a user-centric focus; with client/server computing, all the control res

81、ted with the mainframe—and with the guardians of that single computer. It was not a user-enabling environment.</p><p>　　Peer-to-Peer Computing: Sharing Resources</p><p>　　As you can imagine, acc

82、essing a client/server system was kind of a “hurry up and wait” experience. The server part of the system also created a huge bottleneck. All communications between computers had to go through the server first, however i

83、nefficient that might be.</p><p>　　The obvious need to connect one computer to another without first hitting the server led to the development of peer-to-peer (P2P) computing. P2P computing defines a network

84、 architecture in which each computer has equivalent capabilities and responsibilities. This is in contrast to the traditional client/server network architecture, in which one or more computers are dedicated to serving th

85、e others. (This relationship is sometimes characterized as a master/slave relationship, with the central ser</p><p>　　P2P was an equalizing concept. In the P2P environment, every computer is a client and a s

86、erver; there are no masters and slaves. By recognizing all computers on the network as peers, P2P enables direct exchange of resources and services. There is no need for a central server; because any computer can functio

87、n in that capacity when called on to do so.</p><p>　　P2P was also a decentralizing concept. Control is decentralized, with all computers functioning as equals. Content is also dispersed among the various pee

88、r computers. No centralized server is assigned to host the available resources and services.</p><p>　　Perhaps the most notable implementation of P2P computing is the Internet. Many of today’s users forget (o

89、r never knew) that the Internet was initially conceived, under its original ARPAnet guise, as a peer-to-peer system that would share computing resources across the United States. The various ARPAnet sites—and there weren

90、’t many of them—were connected together not as clients and servers, but as equals.</p><p>　　The P2P nature of the early Internet was best exemplified by the Usenet network. Usenet, which was created back in

91、1979, was a network of computers (accessed via the Internet), each of which hosted the entire contents of the network. Messages were propagated between the peer computers; users connecting to any single Usenet server had

92、 access to all (or substantially all) the messages posted to each individual server. Although the users’ connection to the Usenet server was of the traditional clien</p><p>　　That said, not every part of the

93、 Internet is P2P in nature. With the development of the World Wide Web came a shift away from P2P back to the client/server model. On the web, each website is served up by a group of computers, and sites’ visitors use cl

94、ient software (web browsers) to access it. Almost all content is centralized, all control is centralized, and the clients have no autonomy or control in the process.</p><p>　　Distributed Computing: Providing

95、 More Computing Power</p><p>　　One of the most important subsets of the P2P model is that of distributed computing, where idle PCs across a network or across the Internet are tapped to provide computing powe

96、r for large, processor-intensive projects. It’s a simple concept, all about cycle sharing between multiple computers.</p><p>　　A personal computer, running full-out 24 hours a day, 7 days a week, is capable

97、of tremendous computing power. Most people don’t use their computers 24/7, however, so a good portion of a computer’s resources go unused. Distributed computing uses those resources.</p><p>　　When a computer

98、 is enlisted for a distributed computing project, software is installed on the machine to run various processing activities during those periods when the PC is typically unused. The results of that spare-time processing

99、are periodically uploaded to the distributed computing network, and combined with similar results from other PCs in the project. The result, if enough computers are involved, simulates the processing power of much larger

100、 mainframes and supercomputers—which is neces</p><p>　　For example, genetic research requires vast amounts of computing power. Left to traditional means, it might take years to solve essential mathematical p