By Dr. Mike Johns, NetQoS Product Research Engineer, Netcosm designer
Netcosm started as a bunch of yellow blocks flying between purple blocks. But even at that early stage, the reaction took me by surprise. People from everywhere in the company stopped by to take a look.
The reaction from the Internet at large has also been astounding – but to me, the creator of the software, that reaction has been a bit baffling. What is it about flaming servers, (aside from sheer novelty) that provokes such a response? I believe it is because Netcosm presents network information in a way the brain actually likes to deal with.
There has been explosive growth in the amount of information available to us, and human comprehension is a bandwidth-limited resource. Network engineers are familiar with the general thinking about solutions to bandwidth-oriented problems: either send less information or find more bandwidth.
The fact is, the amount of information processed by the brain is limited by how much information can be crammed through the senses, and the relevance – or signal-to-noise ratio – of what gets through. Sometimes we’re stuck sifting through a lot of noise for the signal we need, so we look for new ways to present it all that simplify the task.
The idea of opening up bigger pipes into the brain doesn’t make a whole lot of sense on the surface. “Upgrading the hardware” is a process that takes millennia. So what little attention has been paid to addressing the problem of increasing the volume of information that the brain can process at one time has become the art of designing the User Interface.
However, typically, today’s business applications tend not to use as much of our natural equipment as is available. Everything tends to get forced through the channels that process text and images. While these are important methods of delivering information, interpreting graphs is a relatively difficult task for the brain to engage in.
But there is a trick we can use. We can process more information in less time by using the brain’s abilities to filter information when it’s delivered in certain ways – specifically, animation and sound. The human brain is exceptionally good at detecting slight anomalies in movement and sound patterns: it helped us eat when we had to kill our own food.
It can take thousands of graphs to convey all of the information shown in 15 minutes worth of Netcosm animation, and most of those graphs would show nothing noteworthy. Instead of deciding how to tell interesting graphs from those that aren’t, however, we can instead present everything and defer to the natural abilities of the user’s attention mechanisms. It is easier to hone in on what is most interesting in such an animation than it is to sort through thousands of graphs.
By presenting, for example, metrics related to speed as an actual rate of moment, and metrics related to volume by scaling onscreen elements, we have a representation that greatly simplifies the work needed to identify and interpret the most important parts of the dataset.
Once “something interesting” is noticed, the full detail of the data can then be examined in whatever form the user prefers.
I really wish I could say that I began work on Netcosm with these principles in mind. Instead, it was, in the beginning, just something I enjoyed watching. But only after realizing that I was becoming acutely aware of the performance of our internal network as a result of developing Netcosm did I begin to believe in its value as a serious tool.
I could watch was going on while sitting idly in my chair, even when primarily engaged in something else. And when an atypical event did begin to unfold, it was immediately apparent. Whether I was watching attentively, talking to someone, or alt-tabbed in another application, the visual and auditory cues immediately grabbed my attention and let it go when things went back to normal — which, as someone researching methods of getting the most important parts of very large datasets examined and understood, was exactly what I hoped would happen.
No comments yet.