Visibility, or how monitoring
serves application performance

As previously said in our post dedicated to WAN optimization stakes, this topic has become inescapable for CIOs. Indeed, wan optimization permits to increase network performance and hence assures a quality of service up to what users demand. Substantial savings also come as a result.

This new press is for us the opportunity to come back on the topic of visibility when applied to a WAN network. Here is explained how visibility is important when WAN optimization is foreseen. e-Qual’s approach of a visibility service on WAN network is described below.

Implicit in the basic understanding of visibility applied on a WAN (Wide Area Network, or extended network) lays the possibility to give CIOs some factual and precise information on what transits on their enterprise network.

When there is no visibility and users complain with problems of network performance, CIOs cannot know what applications are consuming a lot of bandwidth.

It is true that operators sell Internet access and give a whole vision of traffic in terms of volume. But this does not permit to know what applications are the cause of slowness (business traffic? messaging? Internet?). Data agregation adds to this unclarity. Because operators are delivering only a static image of what transits on network. Speed that is volatil and fluctuating, is transformed into averages of fixed data that are not very representative. Thus, WAN remains a black and blur box. And conclusion is always the same: too much data is transiting on network.

The easiest solution is then to increase bandwidth. This leads to additional costs. Such a solution may solve immediately a problem. However this solution is not satisfactory over time. It is only a temporary and pricy solution aiming to reduce complaints from users.

Solutions of metrology we implement to warrant visibility on a WAN, allow us to answer these 3 questions:

Visibility ought to bring lightening on volumetry, in a refined way, keeping maximum data and several levels of granularity. This way, applications and services transiting on WAN may be seen in details. Besides, it is possible to gather applications having the same behaviour in order to obtain a synthetic vision of traffic.

Screening is also very important. Indeed, it is a matter of collecting more subtle information around problems of delays, latency time or lost packets for instance. Mind that our tools allow us to have visibility in real time (less than a minute of lag between visibility and real traffic on network).

Finally, thanks to reporting, it is possible to analyze trends coming from the collected metrology information. Such information can be observed on different levels of granularity: hourly, daily, weekly, monthly, yearly and are available anytime for our customers.

Alarming (warning system) is another advantage of visibility. Analytic indicators allow us to implement warnings as soon as some thresholds are crossed for these indicators.  It is meant for our teams to act in a proactive manner in case of problem.

Some capacity planning functionalities also permit to estimate and plan the required bandwidth for the telecom link leading to the applications used. How behave applications over time is also apprehended through capacity planning.

As seen previously, visibility enlightens WAN. Information transiting on WAN are seen. Parasite traffics (traffics that are not business neither priority flows) as well as their behaviour are visible. In observing all this, it is possible to limit the bandwidth in order to reduce delays and costs related to WAN. It is also possible to adapt analysis considering problems of performance with specific applications.

Does this visibility solution interest you? Are you willing to start a process of WAN optimization? Feel free to contact us. We will answer you very shortly.