Skip to main content

Things to consider for a better internet failover setup

No matter what internet connection or connections you have, they’re each going to have downtime and at times be practically unusable because of performance issues. Based on our customer data, we’ve calculated that ISP connections are providing an average of 93.1% of usable uptime, which translates to an average of 604 hours a year of effective downtime. And with so many important business technologies moving to the cloud, when this amount of downtime, or even a small portion of it occurs, it’s going to be painful and costly.

That’s why having a secondary internet connection that provides failover coverage is such a good idea — and table stakes for many businesses that rely on the internet.

That said, there are a lot of options and things to think about when choosing a second (or third) internet connection and creating the internet failover setup that’s right for your business — whether it’s the first time your business or organization will have a backup internet connection, or you’re looking for a better and more reliable option.

All sorts of things can take an internet connection down  

  • ISP outage
  • Scheduled or unscheduled maintenance
  • Natural disasters
  • Distributed denial-of-service (DDoS) or cyber attacks
  • Human error
  • Hardware problems or failure
  • Power outage
  • Someone cutting a line to your building
  • Spikes in latency or packet loss that make the internet unusable

 

Your primary goal

What do you want out of your additional internet connection and failover method?What’s most important for your business? Is it…
  • Getting your uptime percentage as high as possible?
  • Having the least possible downtime when a natural disaster hits your area?
  •  Improving the performance of a particular cloud technology — like video or VoIP calls, your CRM, or an application that’s specific to your industry or business? 
  • Avoiding interruptions when you fail over from one internet connection to another? For example, making sure VoIP calls or VPN sessions don’t drop. 
  • Something else? Explicitly identifying your main goal or goals will help you make the best decisions for your business and help you explain them to your manager, executive team, or company

Getting the most out of your secondary internet connection(s)

ISP diversity

Even if you have multiple internet connections with a failover option that have so far worked perfectly, you can still have issues. When all your connections are from the same ISP or carrier, they will all experience downtime or serious performance issues when that ISP’s network goes down. When you have ISP diversity, that is, internet connections from different providers, you and your failover setup will have a much better chance of being able to route around issues affecting one ISP or carrier’s network.
 

Last mile diversity

Similar to ISP diversity, it’s also helpful to have physically diverse paths in the “last mile” to your offices or locations. For example, you can combine fiber and cable, DSL and wireless, or T1 and cable so you have more than one method for getting traffic in and out of your site. That way, you don’t have to worry about a construction crew accidentally cutting the line of both your internet connections.
 

The uptime of different connection types

If your business is in an area with a decent variety of ISPs and internet connection types, you might as well pick the connection types that provide an ISP, last mile diversity, and the best shot at maximum uptime. From analyzing the uptime of our 1,700+ customers’ various internet connections, we saw these connection types deliver the most reliable percentage of uptime, in this order: fiber, enterprise fixe wireless, cable, copper, T1/T3, other fixed wireless, DSL, cellular, satellite.
 

Getting the most out of your internet failover setup

Think about outages and performance

Many traditional internet failover options — like dual-WAN firewalls and BGP routers — only jump into action when your primary internet connection completely fails. They don’t have any awareness of network performance metrics for things like packet loss, latency, and jitter that can make the internet practically unusable when they occur, especially when using Zoom, Teams, or other VoIP services. For many businesses, these performance issues are a bigger and more common problem than full-fledged outages. A basic failover setup will be of little help, as they monitor for connectivity failures, not connection health.

 

IP address change

When your primary internet connection falls down on the job and your traffic is moved to your secondary connection, do you want your users’ IP addresses to change or stay the same? For more simple things like email or loading web pages, a change in IP address isn’t a big deal and your users won’t know that your internet was having any issues. However, many cloud- and internet-based applications aren’t so forgiving.

Here are some of the things that can happen when an internet failover changes your IP address:

  • VoIP calls drop
  • VPN sessions disconnect
  • Virtual desktop sessions drop
  • SSH sessions drop
  • Lost data for people editing in electronic health/medical records, CRMs, etc
Most internet failover methods change your IP address(es) when they move your traffic from one internet connection to another. If keeping your IP address(es) the same through any failovers is important, you’ll want to look at options like Bigleaf Networks ora Border Gateway Protocol (BGP) router.
 

Active-active or active-passive configuration

When you have multiple internet connections, your secondary circuit(s) can be passive —just sitting and waiting for your primary connection to go down —or active and sharing the traffic load with your primary connection. Traditional internet failover options have an active-passive configuration where the secondary circuit is there strictly as a backup. This helps you avoid outages, but when your internet connection fails over you will likely have some disruptions and dropped sessions when your IP address or addresses change.With an active-active configuration, both or all your internet connections are actively carrying some of your traffic at any given time. You can even have different types of traffic routed to the connection that’s currently best, for example, the one with the lowest packet loss for your video conferencing platform and the one with the highest throughput for downloads. Additionally, an appropriately configured active-active configuration is unlikely to suffer disruption and dropped sessions when one or the two connections should fail, or suffer high congestion.
 

Bi-directional QoS

Traditional failover options that are a piece of hardware at your locations don’t have any control over your download traffic. This could be fine for your business, but if you’d like to prioritize important traffic that’s particularly susceptible to internet performance issues over bulk downloads, for example, VoIP or video calls, or to be able to route upload and download traffic on different circuits based on the best path, you’ll want a failover option that gives you bi-directional QoS.

 

How much time you have for policies and configurations

Beyond simple failover when one of your internet connections completely goes down, there is a lot that can be included in a failover setup to prioritize and route different types of traffic so that your most important technologies work as well as they can. This can be done through either policies and custom configurations or AI-powered software.

Policies and configurations

With policies and custom configurations, you spell out all the things you think your failover setup will need to know —from telling it how to recognize traffic for your business’ most important applications, to what to do if the packet loss on a circuit crosses a certain threshold. This gives you full manual control, but also takes a lot of time and creates opportunities for human error.

Another thing to keep in mind is that your policies and configurations can only be as good as what your team knows about and has the time to update. For example, when an employee uses a new application they didn’t get from the IT team—a potential customer invites your salesperson to a video conference on a different platform, or the new tool a team is trying out —they won’t get the preferred experience they do with the applications you’ve created manual policies for.

AI-powered software

If you incorporate AI software into your internet failover setup, it can automatically monitor your circuit performance, detect and classify new technologies and traffic types on your network, and route and reroute your traffic to prevent disruptions.Instead of manually creating policies and configurations to try to account for anything that could happen, you can use software that incorporates all the knowledge from the networking experts who created it…the businesses that have already used it.

Adding AI to your internet failover setup is something to seriously consider if you don’t have the time or people to write, test, and debug thousands of lines of policies and configurations, or if the uptime and performance of your cloud-and internet-based technologies is particularly important to your business.

 

Choosing the internet failover setup that's best for your business

Internet failover isn’t one-size-fits-all. What’s right for one business may not make sense or be reliable enough for another, particularly if they have a difference in IT staff resources, budget, and how much their business relies on cloud- and internet-based applications.
 
Here at Bigleaf, we’ve focused our product and support on making it easy for IT people and teams at small and medium-sized businesses to make their internet reliable. If reliable internet connectivity is imperative to ensure that all the cloud- and internet-based applications your business relies on works as needed, we invite you to learn more about Bigleaf and request a demo.