什么是云計算呢？What is the cloud computing?

　　最開始，人們使用算盤。In the beginning, people use the abacus.

　　后來，人們用電腦。Later, people use computers.

　　再后來，人們有了網(wǎng)絡(luò)。Still later, people have "www".

　　再后來，中國人口大爆炸，男女比例男的比女的多3700萬，這三千多萬人沒事干，都去上網(wǎng)。于是服務(wù)器吃不消了。Still later, China's population explosion, the population of men more than women to 37 million, this more than three thousand people have nothing to do but going online. Then the webhosting tired.

于是人們就發(fā)明了NB的技術(shù)，用更好更多的服務(wù)器。So people invented the "NB" technology, with better and more servers.

　　再后來，人更多了，于是服務(wù)器也更多了。Later, more and more people, and more and more severs.

　　但事實上這樣的效果并不好，過度繁重的結(jié)構(gòu)加大了網(wǎng)站設(shè)計和構(gòu)架的難度，而且越是復(fù)雜的系統(tǒng)越是不穩(wěn)定。有可能一個出問題，這樣一個完整的系統(tǒng)就徹底掛掉。如果考慮到系統(tǒng)的崩潰情況，那勢必要引入一個更復(fù)雜的方案來保證不同的服務(wù)器可以做不同的支援。這是一個無解的循環(huán)，大量的計算資源被浪費在無限制的互相糾結(jié)中，很快到了瓶頸。But in fact this effect is not good, over-heavy structure increased the difficulty of site design and architecture, the complexity of the system became more unstable. There may be such a problem, system to completely hang. If you take into account the situation of the collapse of the system, it is bound to introduce a more complex program to ensure that different servers can do a different support. This is a cycle of no solution, a large number of computing resources are wasted on an unlimited intertwined, the soon will be a bottleneck.

　　人們想，那我不用這么亂七八糟復(fù)雜的系統(tǒng)，我上個極其牛逼的服務(wù)器不就好了……而且最牛逼的也還沒制造出來.People think, then I do not use such a mess complex systems, a very fast hardware on my server will be good... ... but the "NB" servers have not yet produced.……

于是人們突然想到了一個好辦法：把所有計算資源集結(jié)起來看成是一個整體(一朵云)，通過并發(fā)使用資源完成操作請求。每個操作請求都可以按照一定的規(guī)則分割成小片段，分發(fā)給不同的機器同時運算，每個機器其實只要做很小的計算286機器都輕松完成的。最后將這些機器的計算結(jié)果整合，輸出給用戶。對用戶看來，他其實根本面對的不是許多機器，而是一個似乎真正存在的計算能力巨牛無比的單個服務(wù)器，比十臺System z10 大型主機揉一起，或是開創(chuàng)了petaflop新紀(jì)元的"拂曉號"與"紅杉"號還要牛。事實上這個服務(wù)器是不存在的，但它擁有著成千上萬臺服務(wù)器的能力。Then people suddenly thought of a good way: gather together all of the computing resources as a whole (a cloud), through the concurrent use of resources to complete the operation request. Request for each operation in accordance with certain rules can be divided into small pieces and distributed to different machines operating simultaneously, each machine in fact do little,but the together is amazing. Finally, integrating the results of these machines, take the result to the user. The user view, he in fact did not face many machines, but seems a single huge cattle immense computing power server real exist , more than ten "System z10" mainframe combine together. In fact, this server does not exist, but it has a capacity of thousands of servers.

　　下面來看實例。See examples below.

　　實際上過程沒這么簡單。哪怕是統(tǒng)計收集資料的過程也會占據(jù)可怕的處理時間。這就將云計算的任務(wù)進(jìn)一步劃分CPU干什么，處理哪個任務(wù)段。這個其實可以由算法安排成自動In fact the process not so simple. Even the statistical data collection process will take a lot of time. This will be further divided cloud CPU doing the task, which task processing segment. In fact, this can be arranged by the algorithm automatically

　　總之，壓榨每一個步驟的潛力，讓一個任務(wù)被服務(wù)器集群們一起上，自然能飛速達(dá)成。別忘了，云計算不是弄個兩三臺質(zhì)低價廉的服務(wù)器就可以達(dá)成的。每一朵云背后都有著一大堆異構(gòu)平臺服務(wù)器，只有很少的IT大佬家可以集齊全套設(shè)備，普通人家沒有、也無需有。In short, the potential for squeezing every step, so that a task by the server cluster together, and naturally reached rapidly. Do not forget, two or three sets of cloud computing is not getting a cheaper quality than low-cost server can achieve. Behind every cloud have a lot of heterogeneous platform server, only a few big companies can collect a full set of IT equipment, ordinary people did not, and need not have.

對于普通用戶來說云計算是一種方便使用的IT資源，就如自來水一樣，擰開龍頭就可以用，而無需在自家后院挖個機井，甚至修一個自來水廠。Cloud computing for the average user is an easy to use IT resources, such as tap water, don't have to dig wells in their backyards, or even build a water plant.