SD-WAN Troubleshooting

Stop the finger-pointing: for effective SD-WAN troubleshooting.

Ricardo Belmar
Jul. 30 2019

Most software-defined wide-area network (SD-WAN) solutions claim to provide tools for SD-WAN troubleshooting. For many solutions though, “troubleshooting” just means isolating faults. Any solution can show you if some component is offline. Try to find out why an application on the WAN isn’t performing the way it should, however, and you’ll likely end up with more questions than answers.

For modern digital businesses, poor application experiences have a direct impact on the bottom line. – on average costing businesses 3% of annual revenue according to EIQ Research. So, before you invest, make sure your SD-WAN can help you spot those types of problems—and do something about them.

Defining SD-WAN Troubleshooting

Detecting faults is now table stakes for SD-WAN solutions. But the problems that take the most time and effort to fix—the ones that can really disrupt your business—demand a more sophisticated approach. Problems like:

  • Important business applications running too slow at some locations
  • Conferencing tools like Skype taking too long to load or never connecting for some users but not others—even at the same site
  • Supply chain applications failing to complete transactions, so needed supplies don’t get ordered
  • Point-of-sale systems suddenly getting hung up at unpredictable times

When users experience problems like these, their first instinct is to blame the network. But is that really the problem? Or is it an application issue? Or maybe some combination of both? Most SD-WAN troubleshooting tools won’t offer much help finding out.

Network engineers look at the fault monitoring screen and everything looks good, so they say it must be an application issue. Application teams don’t see any problems and point a finger right back at the network. Now what do you do?

Building Better SD-WAN Troubleshooting

To provide real help with SD-WAN troubleshooting, your solution should be:

  • Proactive: You shouldn’t have to wait for users to call the helpdesk. If applications on the WAN aren’t performing up to spec, your SD-WAN should recognize the problem and, ideally, do something about it before it affects users.  
  • Automated: Along those lines, your SD-WAN should be able to respond automatically to most problems without human intervention. If there’s an unexpected spike in usage of a critical application, for example, the solution should immediately allocate more resources, without anyone manually forcing it to.  
  • Granular: SD-WAN troubleshooting tools should go much deeper than the overall network level. You should be able to follow the path of application traffic across the WAN port, the site’s LAN connection, all the way down to an individual user’s device. So, if there’s an unexplained slowdown, you can drill down to the level of individual applications and users, and compare problematic sessions to good ones to understand what’s happening.
  • Holistic: To the extent that most SD-WANs offer troubleshooting tools, they’re typically built for networking people. If there’s an application issue, you need to bring in an entirely different set of tools. But what if, for example, a link is running too slow, so an application has to keep retransmitting data? Expect lots of finger-pointing between networking and application teams while they try to figure that out. A better solution will provide multi-layer visibility to isolate complex issues.  

Get a Deeper View into Your Network

With so much happening in a dynamic business, it’s inevitable that problems will crop up on the WAN. But they don’t have to impact your users or your bottom line. Fortunately, the latest generation of SD-WAN solutions are built from the ground up to deliver great application experiences—including the ability to monitor and troubleshoot problems all the way down to the level of individual sessions.

To learn more, visit


Written By

Need updates? Get on our mailing list.