Monitoring the Remote Workforce with Endpoint Agents at AIM Specialty Health

Posted by on August 11, 2016

Yesterday we launched a brand new product and our newest innovation, the Endpoint Agent. Extending the periphery of network intelligence all the way to the end user, the Endpoint Agent provides real user perspective on critical applications and the impact of the underlying network on application performance. In today’s post, we’d like to share the experience of one of our beta customers to demonstrate how Endpoint Agent works in the wild.

Earlier this year Ivan Shepherd, Senior Manager at AIM Specialty Health, a medical benefits provider for specialized services, presented at ThousandEyes Connect where he discussed how his team relied on Cloud and Enterprise Agents to deliver services across an evolving telecommuter network. See the write up of the first half of Ivan’s talk. What we left out from the recap was how AIM Specialty Health uses Endpoint Agents (yes, you read that right!) in their network. In today’s blog post we will dive into the second part of Ivan’s talk on how Endpoint Agents provide yet another vantage point within AIM’s telecommuter network.

Early Adoption

As you can imagine, the Endpoint Agent didn’t happen overnight. Over the past year we have been silently laying the framework by recruiting customers to battle test the Endpoint Agent. AIM Specialty Health was one of our early adopters, employing more than 200 Endpoint Agents within their network to troubleshoot their work-from-home employee performance. If you recollect Ivan’s presentation, he starts the session recounting how he was tasked with architecting a network for specialized medical professionals connecting from remote locations in the U.S., called the Telecommuter program. The Telecommuter program was meant to relax corporate boundaries and allow remote users to be a part of AIM’s taskforce but required the same flawless user experience as campus employees.

Through the evolution of the telecommuter program Ivan’s team was engaged with ThousandEyes, instrumenting both the Cloud and Enterprise Agents, to provide network and application level visibility. But, that wasn’t enough. “Although the Cloud Agents are extremely powerful, they do not cover all locations and permutations of what the local last mile performance is”, says Ivan. And that is where the Endpoint Agent bridged the gap. Understanding work-from-home user performance as it relates to web application performance and network behaviour was critical to AIM Specialty Health, as it directly affected the service provided to their customers. If a remote call center employee was unable to get on the phone and assist their customers, it affects service quality metrics. From a business perspective, it was important to gather persistent per-user data to understand trends and behaviour.

Monitoring Work From Home Employees

Ivan recalls how easy it was to deploy the Endpoint Agents through MSI packages. Since they are installed with a browser-based extension, Endpoint Agents monitor and sample real user HTTP activity to collect both application and network level performance data as opposed to scheduled tests from Cloud or Enterprise Agents.

Ivan then navigates through some of the Endpoint Agent functionality, drawing attention to page completion metrics and says “If our employees cannot reach a web page, then it bears an investigation.” Network packet loss statistics, gathered from all installed Endpoint Agents, as shown in Figure 1, help his team to quickly identify that there could be an underlying user connectivity issue.

Figure-1
Figure 1: Network packet loss spikes to 25% for an Endpoint Agent.

The Path Visualization component of the Endpoint Agent, similar to the existing view for Cloud and Enterprise Agents, provides a hop-by-hop view of the network from the Endpoint Agent to the end domains that are being monitored. This functionality allows Ivan and his team to understand how the remote workforce is looking like at any given time, as shown below in Figure 2. As the network architecture involves VLAN trunking over public transport, path visualization only shows a single hop between the remote user and the gateway server within AIM’s datacenter. Ivan says they are evaluating other alternate network architectures to get past this single hop scenario to maximize insights provided by the Endpoint Agent and gain visibility into the public transit.

Figure-2
Figure 3: Path Visualization provides hop-by-hop view of the network from remote users to monitored domains.

The Endpoint Agent Session Details view, he says, provides beneficial data into the type of user connectivity, be it Ethernet or Wi-Fi. Insight into quality of connection from the end user to the gateway combined with waterfall view of the destination domain helps accurately identify where an issue lies.

What Next?

Ivan and his team are currently evaluating different ways to slice and dice the rich data set they have collected through the Endpoint Agents to help their operations team. “The information is also helping us with selecting new locations to build additional points-of-presence,” says Ivan.

If your organization is riding the wave of remote user workforce combined with migration to the cloud, and you are responsible for ensuring user experience and performance, then the Endpoint Agent might be for you. Join us for a webinar to learn more about the Endpoint Agent and see a live demo. If you are the adventurous type, then dive straight into a trial and see what the Endpoint Agent can do for your network.

Processing...