私有云pk公有云——這場(chǎng)戰(zhàn)爭(zhēng)你贏不了丨雙語(yǔ)

原文作者：Matt Asay

譯者：李哲

關(guān)鍵詞：私有云，公有云

核心提示：私有云還是公有云？這已經(jīng)是老生常談了。本文觀點(diǎn)認(rèn)為，亞馬遜、谷歌、微軟等企業(yè)提供的公有云平臺(tái)，將在與私有云的創(chuàng)新競(jìng)爭(zhēng)中保持完勝。

上周，微軟、亞馬遜、Alphabet（谷歌）的財(cái)報(bào)相繼出臺(tái)，宣示了三家公司云業(yè)務(wù)增長(zhǎng)的強(qiáng)勁勢(shì)頭（增長(zhǎng)率分別為116%、55%和“所有產(chǎn)品線中最高的增長(zhǎng)率”）。然而，事實(shí)還不止如此。

這些公司還宣布，支撐云服務(wù)的基礎(chǔ)設(shè)施建設(shè)是投資重點(diǎn)。投資數(shù)額是其他公司望塵莫及的。那么，這是否意味著你的私有云已經(jīng)淪落成“二等云公民”了？很不幸，事實(shí)基本如此。因?yàn)槟愕膭?chuàng)新速度完全跟不上。

“但是我可以像谷歌一樣運(yùn)營(yíng)啊！”

“服務(wù)器擁護(hù)者”過(guò)去常常聲稱(chēng)，企業(yè)的很多工作不能轉(zhuǎn)入公有云，因?yàn)橛邪踩?、性能或者其他方面的顧慮。隨著這些顧慮的煙消云散，他們又轉(zhuǎn)而論道：“精明的企業(yè)可以更高效地運(yùn)行它們的基礎(chǔ)設(shè)施，而且比使用公有云更經(jīng)濟(jì)?！?/p>

真的是這樣嗎？

對(duì)成百上千的大規(guī)模的服務(wù)器進(jìn)行高效的管理，不是不可能。如果能做到，當(dāng)然很好。根據(jù)451 Research的研究報(bào)告，對(duì)于那些能夠高效管理大規(guī)模服務(wù)器的企業(yè)來(lái)說(shuō)，使用私有云的總體擁有成本（TCO）比使用公有云更低。

但是，這種優(yōu)勢(shì)始終無(wú)法與亞馬遜、谷歌、微軟等公司產(chǎn)生的規(guī)模經(jīng)濟(jì)相提并論。規(guī)模經(jīng)濟(jì)意味著更多更好的硬件、軟件和人才。谷歌宣布，到2017年，云服務(wù)將成為公司最大的投資領(lǐng)域和人數(shù)增長(zhǎng)領(lǐng)域之一，而且將在下一年覆蓋八個(gè)新地區(qū)。微軟已經(jīng)覆蓋了38個(gè)地區(qū)，在不久的將來(lái)還會(huì)進(jìn)軍法國(guó)。

至于亞馬遜？谷歌和微軟只能相形見(jiàn)絀了。

然而，規(guī)模經(jīng)濟(jì)不僅僅意味著花更少的錢(qián)，買(mǎi)更多的硬件。正如Randy Bias所說(shuō)，“創(chuàng)新率和開(kāi)發(fā)率才是這些公有云企業(yè)規(guī)模經(jīng)濟(jì)的真正意義所在。”想知道這句話的真正含義嗎？看看這些公司發(fā)布的開(kāi)源和研究項(xiàng)目吧！這只是它們的數(shù)據(jù)中心研發(fā)投入的冰山一角。

在建立起根深蒂固的“云DNA”之前，不管你的公司有多大，都不可能和公有云同日而語(yǔ)。就像亞馬遜的CFO Brian Olsavsky在公司的電話會(huì)議上強(qiáng)調(diào)的，“我們做這行的時(shí)間比其他人都長(zhǎng)，這些時(shí)間都被用來(lái)完善我們的產(chǎn)品和服務(wù)了?！?/p>

如果，你還沒(méi)被說(shuō)服的話……

當(dāng)然，即使你不再幻想著與公有云的靈活性和創(chuàng)新性相媲美，仍然要把工作流保持在自己的數(shù)據(jù)中心內(nèi)，也是可以理解的。如果是這樣的話，你可以考慮投資被Forrester稱(chēng)為私有云的業(yè)界標(biāo)準(zhǔn)和“安全賭注”的OpenStack。

如果你不是一家電信公司或者沃爾瑪，你可能會(huì)選擇在OpenStack上進(jìn)行開(kāi)發(fā)和測(cè)試工作（dev-and-test workloads）。根據(jù)最近的調(diào)查，在所有使用OpenStack的企業(yè)中，89%服務(wù)的用戶(hù)少于1,000位。沒(méi)錯(cuò)，開(kāi)發(fā)和測(cè)試工作很重要，而且可以鼓舞你在云領(lǐng)域的自信心，為生產(chǎn)工作（production workloads）鋪平道路。但是，OpenStack還遠(yuǎn)沒(méi)有像AWS或微軟Azure一樣擁有如此廣泛的用戶(hù)。

而且它永遠(yuǎn)也不會(huì)。因?yàn)镺penStack服務(wù)的用戶(hù)群在不斷縮小，這些人相信他們需要有自己的基礎(chǔ)設(shè)施，就像上個(gè)世紀(jì)初的制造商相信要有自己的發(fā)電廠一樣。

對(duì)私有數(shù)據(jù)中心的需求是可以理解的，但是，當(dāng)像通用這樣的公司都在宣布要把90%的工作流轉(zhuǎn)入公有云的時(shí)候，我們是否應(yīng)該考慮一下，把默認(rèn)設(shè)置從“私有云”調(diào)整為 “公有云”。

亞馬遜、谷歌、微軟……它們將在創(chuàng)新的競(jìng)爭(zhēng)中保持完勝。

英文原文：

This past week Microsoft, Amazon, and Alphabet (Google) all reported earnings, and all announced impressive growth (116 percent, 55 percent, and “the highest percentage growth of all its product lines,” respectively) in their public cloud businesses. That, however, isn’t all they announced.

These same companies announced major investments into the infrastructure powering their clouds, investments that no single company can hope to match. Does this mean your private cloud is a second-class cloud citizen? Almost definitely. You simply can’t innovate fast enough.

But I can run like Google!

It used to be that server-huggers argued that a wide array of enterprise workloads wouldn’t go to the public cloud due to security, performance, or other concerns. These concerns keep evaporating, leaving apologists to declare, “Well, savvy enterprises can run their infrastructure more efficiently and save money over the public cloud.”

Really?

Sure, it’s possible that you’re awesome at efficiently managing hundreds or thousands of servers at scale. If so, good for you. In a summary of a 451 Research report, Brandon Butler of Network World notes, “Generally speaking, if any organization has the expertise to manage a large number of servers at a high level of utilization, then on-premises, customer-managed private clouds can have a total cost of ownership (TCO) advantage compared to public clouds.”

Ultimately, there’s no way you can keep up with the relentless economies of scale of an Amazon, Google, or Microsoft.?As RedMonk analyst Stephen O’Grady has called out:

[Public cloud providers’] variable costs decrease due to their ability to purchase in larger quantities; their fixed costs are amortized over a higher volume customer base; their relative efficiency can increase as scale drives automation and improved processes; their ability to attract and retain talent increases in proportion to the difficulty of the technical challenges imposed; and so on.

This translates into more and better hardware, software, and people. Google, for example, revealed that “cloud [will] be one of our largest areas of investment and head count growth” in 2017, also stating that the company will add eight new regions next year. Microsoft, for its part, has 38 regions and will add France in the near future. The company also announced it’s spending heavily at the expense of higher profits: “Given the growth opportunity in cloud, we increased our operating expenses by 21 percent … to fund cloud engineering, sales capacity, and developer engagement.”

And Amazon? It dwarfs them all.

But “economies of scale” isn’t merely about the ability to buy more hardware for less. As Randy Bias concludes, “The rate of innovation and development at these public clouds is where the true economies of scale reside.” Want a taste of what he means? Check out the impressive open source and/or research projects these companies keep releasing, a hint of the R&D that goes into their datacenters.

I don’t care how big your company is — until you spend the years building up that institutional cloud DNA, you simply cannot hope to match the public clouds. As Amazon’s CFO Brian Olsavsky stressed on the company’s earnings call, “We’ve been in this business a long time, longer than anyone else, and we’ve used that time to make our products and services better.”

Still not budging

That said, you might not be convinced. Undoubtedly there are valid reasons to keep workloads within your datacenter, even if you can’t hope to keep up with the public clouds’ agility or innovation. As such, you may consider investing in OpenStack, which as Forrester points out?is the default for private clouds and finally a “safe bet.”

Unless you’re a telco or Walmart, however, you’re probably going to stick with dev-and-test workloads on OpenStack, with 89 percent of enterprises recently surveyed using OpenStack to service fewer than 1,000 users. Yes, dev-and-test workloads are important and help breed confidence in the cloud, paving the way for production workloads. But for all the hoopla around OpenStack, it hasn’t come close to finding the broad audience that AWS or Microsoft Azure have.

Nor will it — OpenStack serves the ever-shrinking base of people who believe they need to own their infrastructure, akin to the manufacturers at the turn of the last century that felt they needed to own their power plants.

There are legitimate needs for private datacenters, but when a company like GE announces it’s moving 90 percent of its workloads to the public cloud, it’s probably time to start setting the default to “public cloud” instead of “private cloud.” They’re going to out-innovate you every single time.

生成海報(bào)

免責(zé)聲明：本網(wǎng)站內(nèi)容主要來(lái)自原創(chuàng)、合作伙伴供稿和第三方自媒體作者投稿，凡在本網(wǎng)站出現(xiàn)的信息，均僅供參考。本網(wǎng)站將盡力確保所提供信息的準(zhǔn)確性及可靠性，但不保證有關(guān)資料的準(zhǔn)確性及可靠性，讀者在使用前請(qǐng)進(jìn)一步核實(shí)，并對(duì)任何自主決定的行為負(fù)責(zé)。本網(wǎng)站對(duì)有關(guān)資料所引致的錯(cuò)誤、不確或遺漏，概不負(fù)任何法律責(zé)任。任何單位或個(gè)人認(rèn)為本網(wǎng)站中的網(wǎng)頁(yè)或鏈接內(nèi)容可能涉嫌侵犯其知識(shí)產(chǎn)權(quán)或存在不實(shí)內(nèi)容時(shí)，應(yīng)及時(shí)向本網(wǎng)站提出書(shū)面權(quán)利通知或不實(shí)情況說(shuō)明，并提供身份證明、權(quán)屬證明及詳細(xì)侵權(quán)或不實(shí)情況證明。本網(wǎng)站在收到上述法律文件后，將會(huì)依法盡快聯(lián)系相關(guān)文章源頭核實(shí)，溝通刪除相關(guān)內(nèi)容或斷開(kāi)相關(guān)鏈接。