If you've followed my blogs and research over the past few years, you'll know I've been pushing hard for the network management industry to evolve. My thesis is that the legacy management platforms were developed in an era where IT infrastructure was physical, static, deployed in silos and under the tight control of IT. Today, of course, the whole IT model has been flipped. IT resources are virtual, users own and control endpoints and there are more and more shared resources. Meanwhile, management tools haven't really kept up, which is why so many IT managers I talk to describe scenarios where everything on the dashboard is "green," and yet stuff isn't working.
I walked through the causes for this on a webinar a couple of weeks ago with an up-and-coming network management vendor, ActionPacked! Networks, one of Cisco's many developer partners. In fact, if you want to watch the replay, it can be found here.
During the webinar, we asked a number of polling questions. The first was: "Which of the following is causing the greatest network management challenges?" To no surprise, the top response was "bring your own device," with 55% of the responses. VoIP was selected by 18% of the responders and then video, virtualization and wireless made up 9% each. This response wasn't overly surprising to me, as BYOD is the technology trend that causes IT to feel like they are losing the most control.
The second polling question proved my thesis, or at least that IT individuals agreed with me. A staggering 73% of the respondents answered "yes" to the question: "Do performance problems fall under the radar of your current monitoring solutions". Think about that: Almost three quarters of participants feel that they can't monitor the performance of the applications that drive the company's business. As the IT environment becomes more virtualized, consumerized and mobile, the problem will only become a bigger issue.
The last polling question asked "When performance problems occur, how long do they take to resolve?" The answer to this question shows some magnitude of the impact to the organization. A surprisingly large number of respondents, 40% to be exact, said problems took 4+ hours to resolve. Another 40% stated that problems took 2-4 hours to resolve.
One of the slides I showed in the webinar was the impact of downtime to certain verticals (see below). Taking 4 hours to resolve a problem that causes an application to be down could cost a vertical like a brokerage over $20 million. The top verticals on this chart are heavily dependent on the network, which is why the impact per hour is so high. As more organizations move to network-based compute models like cloud and mobile, the cost per hour of outage will continue to rise.
Source: ZK Research
Closing this gap requires network management tools that can give a holistic view of the infrastructure through the lens of the service. One of the pieces of advice I give IT leaders today is that, before moving forward with cloud, BYOD, video, etc., they need to "know their network". That is, can the IT department visually see what's going on, what the normal baseline is, what's changed and what's a significant deviation from the norm? It's my experience that less than 25% of companies really know their networks, creating a big gap between where organizations are and where they need to be.
Having that knowledge of the network allows IT to be much more proactive and enables the IT department to predict when problems are going to occur and act before the users start complaining. Also, the network can uniquely span the virtual/physical, wired/wireless and cloud/premise boundaries, providing visibility far beyond what can be seen through endpoint agents and other legacy ways of managing the environment. The ultimate goal is to move to a user- or service-centric view of the infrastructure versus legacy management, which gives an infrastructure view, and then the service quality needs to be interpreted.
The sponsor of the webinar, ActionPacked! Networks is a great example of a management vendor that has been able to create a visual, real time view of the infrastructure. There are other up-and-coming management vendors such as Netscout, Riverbed and Gigamon that provide complementary functionality in this new era of IT.
It's clear there's a sea change going on in corporate IT today. Scaling this new world of IT requires many changes, and the one area that often gets forgotten about is the management tools. To the IT individuals reading this, I offer this advice: Be proactive and get out in front of this before it becomes a problem that costs significant money.
