Data Center Journal

VOLUME 36 | FEBRUARY 2015

Issue link: https://cp.revolio.com/i/457819

Contents of this Issue

Navigation

Page 9 of 32

THE DATA CENTER JOURNAL | 7 www.datacenterjournal.com giga-scales for servers and subsequently the data centers in which they are housed. Data center infrastructure moDeLs revisiteD Telecommunications historically has represented an infrastructure model that was "Hardware Defined", meaning that new communication solutions were deployed via a new hardware/soware solution as a package. is package was designed and built with a custom purpose, and as a result, it was a slow process to change. Telecommunication companies built the same basic design of central office (telecom facility) for decades. Conversely, most of the internet companies have built upon a hardware ag- nostic platform where the hardware is all largely the same and inexpensive, only the soware changes. Since soware is able to change much more quickly than the hardware, internet companies have been able to rapidly evolve with technology trends, including social media. is pace of change has driven innovation in data center design, and the need for additional capacity has internet companies building new top-to-bottom designs every year. is is a very simplified way of describing the solution to a challenging problem. e internet companies have built to web scale by employing just-in- time designs that can in turn leverage the latest technologies. ey have provisioned for non-intrusive expansion in addition to designing means for their mechanical and electrical plants to scale. In some cases, they have employed replication at the facility level and leveraged soware applications to make the best use of capacity during rapid growth and spikes in demand. is is the key; the facilities are built to accommodate scaling that is primarily driven by soware upgrades. is, fundamentally, is why some internet based companies have been able to scale at the unpredictable pace of social media. Today's telecommunication com- panies are now converging into an all IP based soware world, where they too can play in the hardware agnostic, rapidly evolving soware world. is transition will enable them to begin competing in the social media and big data races. Big data is the path towards big potential revenues through targeted adver- tising, social engineering, technology dis- coveries, and a whole host of new sciences and means of monetizing information to and of the masses. A 2014 EMC/IDC Digital Universe report indicated, "if the Digital Universe were represented by the memory in a stack of tablets [iPads], in 2013 it would have stretched 2/3 the way to the Moon. By 2020, there would be 6.6 stacks from Earth to the Moon".http://www.emc.com/leader- ship/digital-universe/2014iview/executive- summary.htm ese rates of information growth are what define social media data centers. As a result, they plan differently. ey plan for incredible, largely unpredictable growth. at growth is in communica- tions (data throughput), data storage, and data processing. Because social media is also "socially focused and centered" it must be socially sensitive to managing expectations more so than other types of businesses. As a result, it's not uncommon for social media data centers to plan years in advance for managing unpredictable growth at the rates of expectations. concLusion Social media is a new disruptive means of digital interactions that requires the ability to manage significant scale. e rate that social media data is generated with human to human interactions is un- precedented in the internet. Furthermore, the possible data analytics and use of so- cial media data has even further reaching implications and needs for scale. e telecommunications companies that serve as one of the primary pipe- lines for this social interaction were not originally architected to address this scale. It's also clear that all telecommunication companies are seeing the potential value in further immersing their businesses into the social media data boom. As telecommunication companies strive to meet web-scale demands, they face having to change the conventional mindset and begin scaling more pro- gressively, more like a social media data center would. is is not a trivial or quick transformation. Arguably, the openness of the inter- net has been a primary driver towards the explosion in social media. is openness has enabled everyone to share an abundant amount of information with little to no restrictions. Net Neutrality debates and subsequent regulations are therefore a sig- nificant consideration for the future "scale" of social media and the resulting implica- tions telecommunications networks and social media data centers worldwide. Tomorrow's data demands will be far greater than today's regardless. e subsequent need for both telecommunica- tions companies and internet data centers to clearly identify a scalable data center future has never been more paramount. n the following are some metrics that identify the magnitude of the data growth: Stats of social media and internet sites http://www.intel.com/content/www/us/en/communications/internet- minute-infographic.html Facebook info on photo volumes aug 2014 http://money.cnn.com/2014/08/21/technology/facebook-blu-ray/ : "Facebook users have uploaded more than 400 billion photos, adding an average of more than 350 million each day." twitter volumes http://www.internetlivestats.com/twitter-statistics/#trend https://blog.twitter.com/2013/new-tweets-per-second-record-and-how

Articles in this issue

Links on this page

Archives of this issue

view archives of Data Center Journal - VOLUME 36 | FEBRUARY 2015