While software-defined wide area networking (SD-WAN) offers advantages such as improved security, lower costs, and better control over network traffic, there is often a learning curve when it comes to troubleshooting the network when an application isn’t performing as it should. You need a troubleshooting guide for SD-WAN and network visibility to ensure your applications run uninterrupted.
Your first step is to establish monitoring of your SD-WAN so that you know when the network isn’t performing as it normally does. While monitoring the physical components is relatively simple, the abstraction of the functions can be a bit more complex for monitoring.
Event handling: Your network monitoring begins with examining events, or noteworthy activity on your network. Look for a solution that can send events to a single event-handling system where they can be stored, sorted, and managed. There are budget-friendly solutions available, or more in-depth ones for enhanced analysis. The monitoring tool should automatically generate a ticket or initiate a real-time alert when a critical event has been identified. And all events should appear on a regular summary so they can be handled in a timely way.
Active path testing: Monitoring SD-WAN and network visibility includes determining whether it is delivering an optimized end-to-end service with a high level of reliability. Active path testing determines the paths for a variety of traffic types, and the tests should include the elements of real traffic, including quality-of-service markings, transmission rate, and packet size. One advantage of this type of testing is that it can be conducted overnight or when there is no regular application traffic.
Physical components: SD-WAN is generally run on an x86 system that uses an internal central processing unit (CPU), interfaces, memory, power, and cooling. Any problems with these components should show up on a network events log, but monitoring can also validate how many buffers are used on each path, whether the CPU is saturated at certain times of the workday, and whether power supplies are functioning well. It’s a good idea to verify that the SD-WAN controller will issue an alert or report when there’s a problem with the physical link.
Topology maps: The SD-WAN and network visibility control system should provide physical and logical topology maps. This allows a baseline to be established so that the enterprise can determine the difference between actual and desired dynamic maps.
Troubleshooting SD-WAN and Network Visibility
The guiding principle of your troubleshooting plan is to be methodical. You’ll need to determine if it’s a physical problem, such as a downed interface, a link problem, or a routing or application problem. Work through the following steps:
- Determine the basic functionality of the SD-WAN node, examining the CPU, memory and interface connectivity.
- Evaluate basic interface functionality.
- Virtual private network functionality must be validated.
- Examine integration with overall routing architecture, with no black holes, unreachable subnets, or routing loops.
- Determine a forwarding policy verification to detail how traffic is routed if there is jitter, packet loss, or latency.
Examining SD-WAN and network visibility for troubleshooting can be daunting, but an optimized network with SD-WAN introduces key advantages for enterprise IT. Contact us at Cory Communications to discuss solutions to aid in your network optimization efforts.