Regardless of the size of your company -- Global 2000, medium-sized enterprise or small business -- you might more...
than dislike your WAN provider. You might actually hate them. Perhaps latency is making your voice over IP unreliable, your HTTP connections to software as a service (SaaS) offerings suffer from packet drops, or you're just tired of hearing, "The network is slow." Well, it isn't, and you work tirelessly to ensure that, while getting a crick in your neck waiting on hold with the Internet service provider (ISP). Increasingly, organizations of all sizes are turning to path analysis to ensure connectivity and improve what a WAN carrier has to offer.
If you're a network engineer or IT director for a Global 2000 company, you likely didn't have input into the WAN services you depend on. That contract was negotiated over golf -- ahem, I mean at a conference table -- and bundled hundreds of services into huge contracts. It's a little different for medium-sized businesses, where you may have had direct input as to the carrier selected, often among several carrier options. You might even have budget to run fiber to a better ISP. However, small businesses' IT admins are in many ways in the same boat as their peers at global enterprises; you may only have one real WAN carrier option based on location, and you have to make the best of it.
In all cases, there's nothing better for a good night's sleep, or at least peace to binge-watch Netflix undisturbed, than having the confidence you're not going to be working an emergency ticket when East Coast sales can't get to their SaaS app of choice at 6 a.m. The reality is, you don't have to be the victim of a subpar, or simply distracted, ISP. You have the power to hold them accountable for service-level agreements and ensure they deliver. The trick is to combine the network monitoring discipline you already have with a little discovery and visualization magic to determine what happens to packets after they are dispatched into Border Gateway Protocol hyperspace.
Understanding the unknowable and unfathomable
Network classicists may push back and say the internal path structure of carriers is inherently unknowable, even patently unfathomable. But that's no longer true -- we're beginning to see new network performance monitoring tools able to troubleshoot issues technically beyond our control. The innovation driver is the industry's ever-increasing dependence on critical external services like cloud and SaaS, coupled with an equally increasing number of frustrated network engineers tired of being network apologists for ISP issues.
When you can see into your WAN carrier's network, you arm yourself with details that get issues closed much faster than simply describing anecdotal degraded service effects to yet another series of help desk support levels. It's even more critical when it's transient or the issue occurred hours or days ago. Imagine starting a call with, "Hey, can you take a look at the link between bu-ether16.hstqtx0209w-bcr00.tbone.rr.com and bu-ether.dllstx976iw-bcr00.tbone.rr.com? Latency is abnormally high." You'll be shocked how quickly the issue will be addressed. Even better, how awesome will those details look in the notes of your first trouble ticket update?
Calls come with actionable information: What a concept
These days, and it's surprising to say, I’m grateful for my relationship with our Internet carrier, as a WAN carrier. Very rarely does the Internet, for all practical internal purposes, go down. For example, we once had a backhoe excursion that severed the trunk line for several large customers on campus. Failover then saturated the most commonly shared backup carrier links to the point it affected business. But more often, admins suffer simple internal ISP misconfiguration or routing issues that are relatively easy to fix, but challenging to diagnose. In these cases, the link is up, but specific services, overall performance, or both, are affected.
And the reason things are good for me is that Mike H. at Mega Phone Company gave me his cell number and asked me to call him directly when there's an issue. He's happy to get the call because I can calmly but confidently tell him where the issue is, and I know he'll address it quickly. My calls come with actionable information -- not exasperation and pressure from management -- and often give him early notification of issues that may affect more than just my WAN links. My WAN carrier isn't a problem because we have an information- and metrics-based working relationship that allows us both occasional moments of superstardom, attenuated, of course, with a modest "no problem."
Saving money on carrier services
WANs catch up to MPLS
How to manage service providers