Vice President of Marketing
Now that the cloud is assuming responsibility for many traditional data center functions, enterprises are becoming increasingly dependent on network infrastructure to maintain productivity levels. The problem is, though, that while networks themselves are becoming easier to manage, the traffic they hold is still largely hidden. As Exinda’s Vice President of Marketing, Brendan Reid, points out to IT Business Edge’s Arthur Cole, it’s not easy to tell the difference between legitimate YouTube traffic and cute kitten videos. But with real-time ID, intelligent management and other tools, enterprise executives gain broad insight into how their networks are being used, and how they can be optimized.
Cole: Now that the enterprise is heading toward the cloud, bandwidth management is emerging as a primary concern. What are some of the top challenges organizations face as they seek to optimize available bandwidth?
Reid: The biggest common concern we are seeing across enterprises and educational institutions is the ability to distinguish strategic cloud traffic from recreational Internet traffic. The more the enterprise moves strategic applications to the cloud, the more difficult it will be to decipher what web traffic should be protected and what should be controlled. For example, how can a network manager tell when YouTube is being used by the marketing department for a strategic campaign and when it’s being used to watch cute kitten videos? Without visibility into their traffic, many enterprises will struggle with bandwidth issues.
As a subset of that, we see two more challenges: managing strategic use of social media, and controlling real-time video and collaboration content. The use of social media is growing rapidly in organizations as many look at it as a core part of their marketing strategy. Recognition that the company now uses Facebook, Twitter, YouTube and other social sites for strategic campaigns presents a new problem: How do you build policy and intelligence into the network to better understand the context by which these applications are being used – to decipher when they are a priority and when they are not?
The use of video presents the same issue. Whether it’s a Google Hang Out or YouTube, the use of video in the enterprise is exploding and must be controlled. But what about the strategic collaboration of video and voice, like Microsoft Lync, that needs to be protected? A complex challenge is emerging around these real-time Internet collaboration applications that can’t be optimized in the same way as conventional application traffic.
As a final point, we have to acknowledge that moving more strategic applications to run over the public Internet means that a larger proportion of our business-critical application estate will now rely on an inherently ‘best-effort’ network – far different from MPLS, for example. The burden now falls on technologies such as intelligent bandwidth shaping to address the challenge, as the public network itself is not designed to do so.
Cole: The latest research indicates that only about a third of enterprises can prioritize critical applications on their networks. What are some of the ways in which they can increase both visibility and control?
Reid: As a starting point, enterprises and institutions must adequately segment users, build application tiers and construct SLAs that dictate the desired user experience for all network constituents and traffic. Only once this baseline is in place can we start to intelligently monitor, control and enforce the policies that will govern our business network. An important note is that this exercise will be different depending on the type of company. For example, the strategic applications and key user segments will be vastly different for a university than a manufacturing company. The principles are the same, but the details are vital to build and assure network SLAs.
Once the baseline for visibility and policy is in place, the next step is to identify and filter traffic in real time as it passes in and out of the network. Inspecting traffic at the Layer 7 level is key to differentiating strategic from non-strategic traffic, especially when it is intentionally disguised.
Finally, automating the prioritization of traffic based on contextual variables discovered through the filtering process. Through integration with directories such as Microsoft Active Directory, IT organizations can automate how traffic of a certain type for a certain user at a certain time of day is treated. They can allow X amount of bandwidth to employees for YouTube after 5 p.m. and no bandwidth for YouTube during the day unless they are in, say, the marketing department.
And the secret sauce is to be able to do all this with integrated solutions – otherwise you add to the management headache by having to manage a series of point solutions that aren’t designed to work together.
Cole: Speed and timing are also critical factors when it comes to networking. How can the enterprise boost the real-time factor of traffic and application management?
Reid: Ultimately, user experience is the defining metric for network application SLAs in 2013. Speed is a component of this, but it is not a governing objective in and of itself. User experience has a varying set of implications depending on whether we’re talking about real-time communications traffic for unified communications or file sharing CIFS traffic for SharePoint. In both scenarios, user experience is the defining metric – determining if users are meeting our expectations for productivity and response time.
IT organizations need solutions that can protect real-time traffic quality to make video and voice user experiences reliable, But of course you can’t ‘accelerate’ real-time traffic the way you would for a static file, which can be compressed and optimized. Solutions that can identify what kind of traffic is passing through the network and then apply the right mix of technologies to deliver the best user experience are critical for companies that rely on the network for their key applications.