Data Center Journal

Volume 32 | June 2014

Issue link: https://cp.revolio.com/i/324708

Contents of this Issue

Navigation

Page 28 of 32

26 | THE DATA CENTER JOURNAL www.datacenterjournal.com into something off-putting for their customers, which is the opposite of their objectives. Here, market forces will govern the winners and losers in the use of big data." Despite the clamor against busi- ness collection of user information, the biggest threat comes from the government. e NSA scandal, which has raged for about a year, has been at the heart of such concerns, with the agency following a "collect it all" approach to data. Unfortunately, many technology companies have—whether consensually or otherwise—been con- scripted into this effort, as the Edward Snowden leaks revealed. e privacy issue, regardless of what battles have been won or lost, will remain a top concern, particularly as the U.S. government becomes increasingly authoritarian (through indefinite de- tention and drone attacks in citizens, for example). Big seCurity proBleMs? Storage of massive amounts of data that presumably hold significant value means greater security risks, because the potential rewards for hacking big data stores increases with size. In addition, because anonymous data (and even metadata) carries no guarantee of protecting personal identity, both the digital and physical safety and security of individuals and businesses can be compromised. According to Surdack, "Organi- zations should assume that they have been penetrated, because they have been. e old idiom of building higher and thicker walls (firewalls and such) is as outdated as the Great Wall of China. Cyber criminals are very ad- vanced, very patient and have at their disposal nearly infinite resources, as they take control of literally millions of connected devices to perform their bidding. As a result of this, the offense will always have an advantage over the defense, and organizations will be extremely challenged to keep up with those who would do them harm. Note, simply keeping up with the threat will become more and more difficult." Grossman notes, however, that the threat is not only from the outside. "Some would say that the biggest security concerns are situations like the recent Heartbleed bug or data being compromised by hacking. is may be high profile at the moment, but it's not the biggest concern. e biggest concern, judging from what we are seeing, will be the accidental deletion or removal of information by unauthorized people who are actu- ally inside the organization. ese people are oen well intentioned, but they are asked to do jobs and are given responsibilities that they're not equipped handle. is opens the door to accidental errors resulting in data loss." Lost data, particularly if it is at all selective, can drastically change the results of an analysis. For instance, ef- forts in the E.U. to establish a "right to be forgotten" means that individuals and companies could all but rewrite their electronic histories. In the same way, lost information in the context of big data analytics could result in erro- neous inferences or other problems— with potentially deadly consequences when governments get involved. Of course, security concerns go well beyond the province of big data, so whatever threats arise in the broader IT context are also likely to affect big data specifically. Big future or Big DisappointMent? With so much hype surrounding big data analytics, the future is liable to be at least a mild disappointment. Assuming the technology garners and maintains a prominent position in business, however, a number of concerns will remain, including secu- rity and privacy, as well as propriety: what should and shouldn't companies, governments and even individuals do with all the data out there (much of which is publically available)? e future could thus go any number of ways with regard to big data. Chad Meley sees the Internet of ings as one area where big data has a potential role. "We are still in the early days of machine-generated sen- sor data collecting a meted cadence of raw data from anything. Some applica- tions are becoming broadly under- stood, such as mechanical engines sending information that can be used to analyze and detect the need for pre- emptive maintenance. Nest is a great story where a connected thermostat in your house is monitoring your heating and cooling habits in conjunction with your cell phone to understand your location and if you are inbound to or outbound from your house, all leading to machine learning to optimize your comfort and bank account." Surdack sees huge value coming from information, rivaling the changes that Industrial Revolution brought. "Here now we are truly entering the Information Revolution, where data, information and insight are the source of wealth and power. Again, this transformation will shake our society to its core. Only this time, it will occur over a decade or two, rather than a century." But big data skeptics and others who are more cautious about the future of new technologies might express a more reserved opinion. Although this particular technology may have potential, nothing is proven until it is proven. e next logical development in transportation—the flying car—still hasn't arrived. And ultimately, although information is certainly helpful, it can never replace physical production as the true under- lying value in an economy. Unless, of course, we wish to declare, "Let them eat big data." n Critical Power Monitoring for Mission Critical Success. Introducing the new STARLINE ® Critical Power Monitor (CPM). You want the best for your data center. When it comes to power distribution, STARLINE gives you the best in efficiency, reliability, and customization. Now, with the STARLINE CPM, you can integrate revenue grade power monitoring for both end feed and branch circuits, to fully understand and manage your data center's power usage. To learn more about how our CPM can help your data center or other mission critical installation, please visit www.StarlinePower.com. The STARLINE CPM also features: • Revenue grade metering functionality • Monitors multi-circuit configurations • End feed and branch circuit models • Display and non-display versions available • Wired Ethernet and serial communications are standard, optional 802.11n Wi-Fi connectivity • Ability to use both Modbus and SNMP protocols simultaneously 3022_uec_dcjournal_two_page_c_Layout 1 5/15/14 10:21 AM Page 2

Articles in this issue

Links on this page

Archives of this issue

view archives of Data Center Journal - Volume 32 | June 2014