Three Reasons to Upgrade Your Network Backbone

Businesses that are looking to upgrade their network backbone often come to us when they’re experiencing network quality issues.  These issues typically fall into three categories: 1) tired of dealing with their traditional telco, 2) experiencing recurring outages, or 3) wanting to simplify their network backbone. Let’s dive into it…

1. Tired of Traditional Telcos

Many of the customers I speak with are comfortable trusting what they know. They prefer having a Layer 1 Nailed-Up Connection (NUC) between their sites, and if a NUC fails, at least they know what’s failing and can reroute their traffic over an alternate route of their choosing, if they have one. Then they can open up a trouble ticket with their telco provider and hope this Traditional Telco can find the issue and fix it.  Sometimes customers, after being hit with an outage, will order a diverse connection, either with the same traditional telco or a different one. They end up paying twice because they cannot trust that the original NUC will stay up. Most customers are not necessarily familiar with how Layer 2 (the data layer) works, or they purchased an older Ethernet service a long time ago and got burned because no one could tell when the service went down until the customer noticed it. If a Layer 2 connection fails in a modern software-defined network (SDN), their data can take an alternate route automatically, usually within 50ms, eliminating the need to add a diverse NUC.  If they have their BFD (Bi-Directional Forwarding Detection) timers set up appropriately, customers never even know that they have been rerouted across the SDN platform. 

Increasingly, however, a new generation of network engineers and architects has grown to expect an SDN that emulates the cloud, or Network as a Service (NaaS). They’re tired of calling their carrier and waiting months for a NUC to be installed. In one case, we had a customer who had been waiting over 12 months for diverse services to be installed. 

These network teams, often at cloud-native companies, expect to get connectivity up and running quickly because cloud environments are spun up and down on-demand, and mission-critical applications need to be running 24/7 to maximize revenue. They’ve come to expect that a network service provider can direct traffic in an efficient manner across many different alternate paths between endpoints right away in the case of an impacted service. See below for an example of the many routes packets could take between Chicago and Ashburn. 

Alternate route paths between Chicago and ashburn

2. Recurring Outages

Customers who are less accustomed to using SDNs and are more comfortable dealing with their carriers are not very happy about frequent or chronic outages. 

When you’re reliant on NUCs between sites, those outages can be caused by all kinds of unpredictable reasons: a gopher chewing through a bundle of fibers, inclement weather which causes a mudslide which takes out a fiber run, or human error during routine maintenance.  This is where a NaaS platform can provide fast failover (50ms) to alternate routes so that your services aren’t completely taken down when there’s a problem with a physical Layer 1 network service offered by a single carrier.

Some customers have decided to simplify their network and reduce the number of connections they have to manage by using our Cloud Router (CR), which can provide fully meshed connections to six sites with only six Cloud Router Connections (CRCs).

For the customer who was waiting over 12 months for diverse services from their traditional telco, we simply turned up dual connections for them in the locations they needed (from CHI to SEA) and provided them with LOA/CFAs for each end of those services in minutes. Their services were up and running when the cross connects were completed in a matter of days, instead of over a year.

3. A Need to Simplify the Network

If you have X number of sites and want fully meshed connections to each site, the formula is #NUC = (X(X-1))/2 where #NUC is the number of Nailed-Up Connections required to have a fully meshed network between “X” number of sites.  For example, if you have six sites (6(6-1))/2 = 15 NUCs for a fully meshed network. If you want a fully diverse, fully meshed network, multiply that by two for a total of 30 NUCs. 

Some customers have decided to simplify their network and reduce the number of connections they have to manage by using our Cloud Router (CR), which can provide fully meshed connections to six sites with only six Cloud Router Connections (CRCs). For a fully diverse, fully meshed network for six sites, it would only take 12 CRCs. NOTE: Make sure to take into account the costs of all the cross-connects for those NUCs as well.


fully meshed network connecting Six sites with traditional telco NUCS


fully meshed network connecting six sites with packetfabric 100G cloud router

Upgrading Your Network Backbone?

Are you having network quality issues? Are you looking to upgrade or replace parts of your network backbone? Talk to one of our sales engineers and let us know how we can help you.