译见| 大数据摘要：设置你的幸福的小工具

36大数据专稿，拒绝转载！

大数据的来源不是谷歌也不是IBM，而是在1970年的智利政府的努力转移到社会主义的过程中，这是我们从最新的NEW YORKER杂志（The Money Issue 一文）获悉的。

当时全国具有马克思主义倾向的领导者，萨尔瓦多•阿连德，将国家重点产业国有化，所以他自然想建立一个“超现代信息系统”，这个系统可以向政府官员显示全国的工厂如何生产的以及整个民众是如何的快乐，而且一切都是实时的！

纽约客作家叶夫根尼•莫罗佐夫详细介绍了Cybersyn项目的许多功能——一些实际的，有的只是规划，它预测了我们的大数据驱动、云计算、超连接以及当今世界的物联网。

该系统的运营中心看起来像房间大小的商业智能仪表板。四个屏幕可以显示上百张图片和数字，以及有关工厂生产的统计信息，有益于用向上和向下箭头去总结。该系统甚至可以做一个排序的预测分析。

如果你想知道早期系统是如何以一个可视化的方式呈现这些数据，答案是，手工制作。图形艺术家被聘请来更新屏幕。

在真正云的时代，每家工厂将用相关的数据满足这种“指挥中心”，就好比是有多少供求可以满足，或者工厂生产某种东西的当前效率是怎样的。

Cybersyn甚至预测到物联网。该计划的特点是，客厅般的工具，可以让每一个公民通过在极其不幸福和非常幸福之间的仪表盘上拨号来设定他或她的幸福水准。这种幸福数据也将通过电视无线电波被返回到中央规划，在那里会产生一个国家的幸福指数。

更广泛地说，虽然，今天的大数据系统和Cybersyn惊人地相似，因为彼此都企图“收集来自尽可能多的相关数据，对这些数据进行实时分析，并根据当前的情况做出最优决策，而不是依据一些理想化投影，”莫罗佐夫表示。

尽管所有这些大数据就在阿连德的指边，他也没能抵挡一（CIA赞助的）政变，或政*后他自己的神秘死亡。

至于Cybersyn？接下来的领导者，臭名昭著的独裁者奥古斯托•皮诺切特，不能容忍中央的规划，宁愿让经济在自由市场的环境下发展，这个疯狂的有先见之明的系统也就销声匿迹了。

对于New Relic，大数据已经非常，非常好了。

“我们看到了很多有兴趣的人想用数据来更好地经营他们的业务，”，New Relic的产品高级副总裁吉姆Gochee，在接受电话采访时告诉我们。

New Relic最初给银行提供应用程序性能管理（APM）服务，它给开发人员和管理员提供了对于他们的应用如何运行的有价值的见解。

由于该服务要求用户在他们想监控的应用程序上布置一个数据采集软件，对提供数据分析服务的公司来说，这是一个自然的延伸。

今年早些时候，该公司推出了一项名为Insights的数据分析服务，这项服务建立在APM的基础设施上。该公司没有把Insights搭配给管理员，而是给了业务线的经理。

这种方法似乎已见成效。本周，在New Relic的年度用户大会上，FutureStack，该公司鼓吹Insights现在每月可以收集2.1万亿系统事件。

这是一些大数据。而且它变得越来越大。据该公司计算，Insights的用户每天查询触摸340000亿存储系统。

“给大家介绍一下，整个Twitter的数据流还不到Insights每天收集数据的百分之一，”Gochee说。

就像New Relic，通过采用Splunk的搜索引擎设置在堆积如山的机器数据，Splunk的最初帮助管理员和开发人员监控系统性能和解决问题取得了成功。而像New Relic的，Splunk的扩展了它的技术来给商业领袖提供大数据分析工具。

本周，Splunk深入到大数据的领域，并在Hadoop和Amazon的简单存储服务平台上给数据集提供连接。 “有机器数据的地方就是我们进军的方向，”Splunk的首席技术官托德•帕帕约安努说，他在Splunk自己的用户大会现场接受记者采访时被SiliconAngle抓拍到。

跨过商业智能空间，开源BI软件供应商Pentaho的肯定牢记大数据记，当它开发出最新的商用版本的同名开源BI套件，Pentaho的5.2版本已经发行。

在其他配备之间，一个用于Hadoop和其他数据源之间来回移动数据拖放界面得以更新，英国的Computerweekly报告。

大家对Hadoop都不陌生，Teradata的不断在它的数据仓库系统和Apache的数据处理平台之间搭建桥梁。周四，该公司宣布与Cloudera的合作伙伴关系，这使我们怀疑Teradata的统一数据架构将如何适应与Cloudera的企业数据中心。关键的一点事，我们可以把所有的都放在数据湖中。大数据是最时髦的词汇笑话。

不要忘了，浩繁的部分数据的欣赏者，下周是New York Strata+ Hadoop Worid，今年这将震撼cavernously-large Javits中心。

只是从我们从厂商得来的需求的情况来看，这次会议应产生很多观点，或至少有一些讨论。新的机器学习系统？集线器或湖？快速的数据摄取系统？大的云供应商的新闻？转动你的快乐转盘，回来把幸福找出来。

英语原文：

Big data’s origins lie not with Google or even IBM, but in a 1970’s Chilean government effort to move to socialism, so we learned from the latest New Yorker (“The Money Issue”).

The country’s Marxist-leaning leader of the time, Salvador Allende, was nationalizing the country’s key industries, and so naturally he wanted to build a “hyper-modern information system” that could show government officials how productive the country’s factories were, and how happy the entire populace was—all in real time!

New Yorker writer Evgeny Morozov details how the Project Cybersyn’s many features—some actual and some just planned—anticipated our big data-driven, cloud-computing, hyper-connected, Internet-of-many-things world of today.

The system’s operation center sounds like nothing so much as a room-sized business intelligence dashboard. Four screens could show hundreds of pictures and figures, as well as statistical information about factory production, helpfully summarized with up and down arrows. The system could even do predictive analytics of a sort.

If you’re wondering how this early system could present all this data in a visual fashion, the answer is, by hand. Graphics artists were hired to update the screens.

In true cloud fashion, each factory would feed this “command center” with pertinent data, such as how many supplies were at hand, or what the current rate of production of whatever the factory was producing.

Cybersyn even predated the Internet of Things. One planned feature was a living room gadget that allowed each citizen to indicate his or her own level of happiness by way of a dial that ranged from extreme unhappiness to complete bliss. This happiness data would also then be returned to central planning, by way of the television airwaves, where it would tallied to produce a national happiness index.

More broadly though, today’s big data systems strikingly resemble Cybersyn in that both attempt to “collect as much relevant data from as many sources as possible, analyze them in real time, and make an optimal decision based on the current circumstances rather than on some idealized projection,” Morozov observed.

Despite all this big data at Allende’s fingertips, he failed to fend off a (CIA-sponsored) coup d’état, or his own mysterious death shortly thereafter.

As for Cybersyn? The next leader, notorious dictator Augusto Pinochet, had no truck with central planning, preferring to leave economic progress in the hands of a free market, and so died the wildly prescient system.

Big data has been very, very good for New Relic.

“We’re seeing a lot of interest in people wanting to use data to run their business better,” Jim Gochee, senior vice president of products at New Relic, told us in a telephone interview.

New Relic originally made its bank on an application performance management (APM) service, which provides developers and administrators with valuable insights into how well (or not) their applications are running.

Since the service required that users place a data collecting software agent on all the apps they wanted to monitor, it was a natural extension for the company to also offer data analysis services.

Earlier this year, the company launched a data analysis service called Insights that builds on this APM infrastructure. The company marketed Insights not to administrators but to line-of-business managers.

The approach has seemingly paid off. This week, at New Relic’s annual user conference, FutureStack, the company trumpeted that Insights now collects 2.1 trillion system events every month.

That’s some big data. And it gets bigger. Insight user queries touch 34 trillion stored system events a day, the company calculates.

“To give you an idea of the scale of this, the entire Twitter stream is less than one percent of what we’re inserting into Insights every day,” Gochee said.

Much like New Relic, Splunk originally found its success helping administrators and developers monitor system performance and troubleshoot problems, by using Splunk’s search engine set loose upon a mountain of machine data. And like New Relic, Splunk expanded its technology to offer big data analysis tools for the business leader.

This week, Splunk dove deeper into the big data waters, offering connectivity to data sets in Hadoop and Amazon’s Simple Storage Service. “Anywhere there’s machine data, that’s where we go,” said Splunk Chief Technology Officer Todd Papaioannou, during a live interview caught by SiliconAngle during Splunk’s own user conference.

Striding over from the business intelligence space, open source BI-software vendor Pentaho certainly kept big data in mind when it developed the latest commercial version of its eponymous open source BI suite—Pentaho Version 5.2 is out now.

The updated bits come with, among other goodies, a drag and drop interface for moving data back and forth between Hadoop and another data source, reports the U.K’s Computerweekly.

No stranger to Hadoop, Teradata continues to bridge its data warehouse systems to Apache data processing platform. Thursday, the company announced its partnership with Cloudera, leaving us to wonder how Teradata’s Unified Data Architecture will fit with Cloudera’s envisioned Enterprise Data Hub. Call in Pivotal, and we could drop it all in a data lake. Big data buzzphrase joke. You laugh here.

Don’t forget, appreciators of data in voluminous portions, next week is the New York Strata + Hadoop World, which is rocking the cavernously-large Javits Center this year.

Simply by judging from the meeting requests we’ve been getting from vendors, the conference should generate a lot of insights, or least some buzz. New machine learning systems? Hub or Lake? Fast data ingestion systems? News from a big cloud vendor? Turn your happiness dial to bliss and come back to find out.

本文由36大数据合作伙伴北理大数据教育 翻译，拒绝任何转载！