For years, level 1 help-desk agents at regional service provider EATEL were flying blind in support calls about...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
degraded or slow service. They would spend half an hour troubleshooting problems by going down a checklist: Restart the computer, reset the router, reset the modem.
Often, it was the customer's own activity to blame: BitTorrent was a common culprit for slow speeds, as was hardware hijacked by botnets.
"Worst-case scenario is if we're not able to resolve it over the phone. We dispatch a technician," said Harris Miller, data engineering manager with the Gonzales, Louisiana-based service provider. Even then, the on-site technician might miss the issue if the user were not downloading BitTorrent files at the time. And those fruitless trips were expensive.
Last December, EATEL found a solution to this problem when it began testing Xangati's Virtual Task Manager, an update to its popular Rapid Problem Identification (RPI) appliance. Miller said that in the few months they have been using it, service trips have been cut by 25%, and call times have been reduced as well.
Virtual Task Manager makes NetFlow monitoring available to level 1 help-desk agents through a simplified, browser-based application.
NetFlow is Cisco's widely used network protocol that captures traffic information, like the source and destination of IP packets, as well as their protocol.
Xangati's appliance measures how much bandwidth applications consume and then identifies them by protocol, whether it's a movie download or a gaming session.
"Even if it can't categorize the traffic, [Virtual Task Manager] lets us dig in a little deeper," Miller said. Xangati also lets a wide variety of employees do that digging without extensive training.
"We had more sophisticated tools [that] our IP engineers could use … but nothing scaled down to where a support rep could effectively use it," he said.
EATEL already owned one network visibility appliance, but Miller said promises for an upgraded dashboard for that appliance were never fulfilled. The ability to extend monitoring capabilities to the help-desk agents while giving the network engineers easier tools to work with made the purchase a solid investment, according to Miller.
"The challenge of identifying what the problem is has really changed dramatically [in the network]," said Jim Metzler, vice president of Sanibel, Fla.-based consultancy Ashton, Metzler & Associates. Service problems aren't always caused by network issues, such as a dead router or mis-configured switch. Now users are causing networking at the endpoint with bandwidth-greedy applications such as file sharing.
"These days, if an application is degrading, almost anything could be the cause of it," Metzler said. Properly configured NetFlow monitoring can give a clearer idea of what is going on.
Xangati's focus on rapid problem identification is well suited to this challenge, he said, particularly in cutting call times and getting everyone back to work more quickly.
Before jumping into a solution, however, organizations should have a clear idea of what they want various support levels to do, Metzler said. For many companies, giving level 1 help-desk agents deeper diagnostic data could be a great cost saver. Others might want to reserve those functions for higher-level support, keeping level 1 focused on properly routing trouble tickets.
Another concern is a potential privacy backlash, Miller said, something EATEL takes very seriously.
"We ask for permission to run this connection when they call up," he said. "We tell them we're only looking at headers, not payload." NetFlow can identify that a customer is sending email, for example, but not that the email is to Aunt Bessy about her secret family clambake recipe.
So far, Miller said, these explanations have been enough to stave off concerns. EATEL also has the Xangati appliance alert the provider to irregular connection patterns, which could be a virus infection, for example. In that case, a help-desk agent actually calls the customer to alert him to the problem.
Miller said customers who had been accustomed to sitting online for hours have reacted very positively to the strategy and appreciate a heads-up if their connection has been compromised.
Beyond the help desk, higher-level network engineers also make use of the appliance, Miller said, using it to spot trends and potential trouble spots before complaints flood the call centers.
Metzler said that multi-purpose devices were becoming more common, as customers demand more for their money. He said NetScout and NetQoS were both seeking to broaden the usefulness of their NetFlow monitoring applications, but they were not pushing the help desk angle like Xangati.
"I think there is also just the interest in having a common view of what's going on. Why have multiple views of the universe?" Metzler said. Employees across the networking department view identical data and diagnostics, communication improves, and the chance of conflicting reports is reduced.
Have something to say to this story's author or have a subject you want covered in the future? Email Michael Morisy at firstname.lastname@example.org