Inside Track - getting to know the vendors

Ray Smyth discusses network intelligence, its meaning and its value with Mohit Lad, co-founder and CEO of ThousandEyes

The term network intelligence conveys many things, but applied to the wholesale collection of network related data, it is unintelligent and causes data overload. Then there are borderless networks, another term hijacked by marketing. While networks are borderless, it's more important to understand that they are comprised of multiple elements including LAN, WAN, data centre, cloud and internet. This multiplicity creates a network topology that is highly complex and frequently changing. In such networks, traditional, defined network, SNMP monitoring is ineffective.

I asked Mohit about the data that ThousandEyes generate and its relevance. "We only collect data concerning services the customer cares about: for example, applications, events, location. Starting with an application view, you can identify the devices that form the underlying delivery network and monitor them in detail." Remember that routes and paths change quickly, especially when the internet is included. Mohit adds, "with fine-tuning and visual display, it is easy to focus on the things that matter at a given point, and update in real time, as the topology changes."

It doesn't stop with reporting relevant data, which is passed through what Mohit describes as three lenses. "First is the assessment of critical application health, then fingerprinting to determine root cause. Completing these steps correctly creates essential data outcomes and remediation can be executed precisely and with confidence." Mohit adds that, "This measured approach allows you to involve the correct people in the remediation process, avoiding wasteful finger pointing and guesswork." By adding problem categorisation, a defined remediation workflow, using in-house and contracted resources, can confidently operate.

At the heart of ThousandEyes is the technology led business model. "Born in the cloud companies tend to take more risks, are fast to leverage new technologies, and push internet limits. In contrast, legacy companies are unsure of the internet and operate a more static network topology," according to Mohit. It would be wrong to categorise ThousandEyes as enterprise technology: it is a business tool for the organisation whose existence and operation is network dependent. Mohit points out, "Even if you are not using, for example, AWS, your applications will be making API program calls to AWS hosted services. The network topology that facilitates this could be operationally critical."

Digital transformation and digitally native organisations create the challenge that ThousandEyes meets. It's not the right choice for all organisations and Mohit says that one of the biggest barriers they face is IT staff overwhelmed by too many disparate network tools. Mohit was clear, "We don't replace existing network management tools because what we do is different." This seems fair, but I would suggest that once ThousandEyes is deployed, a detailed review of network tools would cause some tool rationalisation.

For such a young company ThousandEyes, appropriately enough, does not lack vision. "Organisations that are succeeding are network centric: the network drives their ability." This is refreshing, as many vendors have produced solutions that de-skill aspects of networking, and this will likely change as legacy organisations decrease in number. "The new business model created by the digital opportunity produces the need for a new network skill set which extends beyond the network team." Perhaps it is only a matter of time until the born-digital organisations appoint a Chief Network Officer to their board…

To close, I asked about the future. Mohit explained that there were two priorities: "To collect as much and as broad a range of network data as possible, and then to develop active intelligence to make sense of available data, collectively, and to create a holistic understanding of the network, its applications and users, through intelligence algorithms and heuristics. This will improve coverage and accuracy for our customers." NC