We all know the story that a chain is only as strong as its weakest link. In the IT world, that weak link could well be the network, since Big Data solutions providers are focusing on servers, databases, applications and even data center utilities-often at the expense of networks.
I saw an example of this last week-at a very small company of 20 employees (with 120 remote user-agents). The company had opted to replace its DSL/TI-based communications with an expanded one gigabit per second, dedicated fiber optics line. Projects like this are not "rocket science," but what made it exceptional was the fact that a company this small was making such a substantial investment.
What Big Data opportunities on the Internet or from sources outside of corporate walls are companies seeing to justify this investment in networks?
More incoming business intelligence from machine-collected data
United Oil Company began a network rebuilding project by replacing satellite DSL communications for nearly 200 service stations throughout Southern California with 4G. During the project, the company took a closer look at the point of sales (POS) information coming in from service stations as customer payment transactions were being executed-and it found a lot more information in the form of Big Data for these transactions that it could collect and then warehouse at headquarters for business analytics.
Rutgers University and Xerox discovered the value of a cloud-based HPC (high performance computing) network for enterprise customers that are just getting their feet wet with probing Big Data for answers on markets, products and strategies.
At an Internet2 conference that took place last week, there was a definite focus on broadening network bandwidth and improving collaboration through network enhancements that would facilitate better videoconferencing and other Big Data uses.
Activities like these will create larger network pipelines for Big Data-and while not every capability is fully commercialized yet, companies are sensing that they will be. This anticipation is a major driver for network expansion in 2013.
Meanwhile, there are several practical steps related to networks that organization should be taking now to ensure the successful transport of Big Data.
These steps include:
Making network planning part of all Big Data projects
You'd be surprised at the number of companies that put servers, databases, applications and even data center operations all on the Big Data discussion agenda whenever a Big Data project comes up-but forget about the network. Network capacity and sizing should always be an integral part of any Big Data planning exercise.
Thinking about security and data ownership
For companies opting to use cloud-resident Big Data processing and storage, there should always be upfront discussions with vendors on who "owns" the data and on who has the right to use (or reuse) the Big Data mining formulas and algorithms that are developed.
Plugging your network into your Big Data roadmaps
Discussing the network and what it can and cannot do should not be limited to active Big Data projects. The network should also be a major element of any Big Data roadmap that defines the future. As part of this effort, network specialists should be keeping their ears to the ground with respect to new network-enhancing technologies that are likely to be commercially available in the future, and that can be directed into the company's Big Data initiatives.
Mary E. Shacklett is president of Transworld Data, a technology research and market development firm. Prior to founding the company, Mary was Senior Vice President of Marketing and Technology at TCCU, Inc., a financial services firm; Vice President of Product Research and Software Development for Summit Information Systems, a computer software company; and Vice President of Strategic Planning and Technology at FSI International, a multinational manufacturing company in the semiconductor industry. Mary is a keynote speaker and has more than 1,000 articles, research studies, and technology publications in print.