You are here
Do Network Monitoring Solutions Help or Hinder
Considering the very large investments made by businesses and service providers in network monitoring solutions, why is poor service quality still the number one complaint from online customers? Have the customer demands for the online experience changed or is the problem one that is not easily addressed with network monitoring solutions?
Monitoring solutions come in two distinct forms, 'passive monitoring' and 'active monitoring'. Passive monitoring is one of observation; this is a bit like standing on the side of the road to assess traffic levels, i.e. utilization. Active monitoring is one of actual measurement; this is a bit like actually driving on the road to assess performance, i.e. 'experience'.
The problem is not that monitoring fails in its actual task, on the contrary most monitoring solutions are used to collect a wealth of 'monitored information' 24 by 7 365 days per year. The problem arises when asking the question, 'does monitoring deliver the information that allows the provider to proactively identify and eliminate problems that impact the customer experience?' Because no matter which way you slice or dice the problem, to the customer, the online experience is everything.
Monitoring passively is useful but flawed in its goal to ensure the customer experience. First, passive monitoring does not incorporate the user experience and second, passive monitoring is isolated to individual network components. Because of this, the assumption that the passive measurements provide the business with key performance indicators that correlate to customer experience is largely invalid. This is a bit like standing at a junction of a major highway that is several hundred miles in length and being asked to report on the 'driving experience' for users of the highway end-to-end. Such an assessment is obviously impossible and to have any chance of delivering a workable solution would demand that the highway is monitored at each and every intersection.
Monitoring 'actively' instead of 'passively' imitates the user experience synthetically. In other words a better way to understand the drivers experience on the highway would be to actually drive the highway end-to-end and then report your experience in detail. In this capacity an 'active test' solution delivers a method for measuring a wide range of application services, so that a provider can assess whether the services being delivered meet the customer's service expectation or level agreement (SLA actual or implied). Accuracy would be paramount for the results to be of value. In particular, the accuracy and suitability of the test process must align with the application used by the customer or it is a waste of time. To make a point, there would be no value in using a motorcycle when the customer drives a bus.
Given the better fit of active testing solutions to understanding the customer experience, it is not surprising that the demand for active 'experience' testing solutions is rapidly growing. But can 'active' network monitoring solutions deliver?
If you drive along a highway to test the highway and the time taken is too long, then the test report must document all the delay events that occurred during the test process. This level of detail is needed if the delay events are to be resolved and acceptable service restored.
The dilemma is that network active testing solutions can easily measure time (experience) but cannot easily detect or measure cause because visibility to the cause events is not available. Knowing that latency is poor without identifying its cause does not deliver a means for resolution.
The explosive growth of smart devices, and more importantly the growth of applications that are acutely time dependent, such as video and VoIP, led Visualware to pioneer a custom testing platform specifically to measure and report the quality and timely delivery of data.
A custom platform delivers the ability to provide a secure solution, void of a commercial operating system, that is able to expose the test process to the otherwise hidden network problem events. This approach improves accuracy, identifies root cause(s) and operates safely in the public domain.