What the Rogers Outage Taught Canadian Businesses About Internet Redundancy

Most businesses don’t spend much time thinking about their internet connection when it is working. 

When it isn’t, everyone notices. 

The Rogers outage in July 2022 was an unusually large example, but it made a fairly basic point for Canadian businesses: a lot more than web browsing now depends on connectivity. Phones, Teams, payment processing, cloud applications, remote access and customer service can all be affected at the same time. 

With the outage back in the news this summer as the CRTC released its independent review, it is a good time to look at a question we still get from customers: what actually counts as a backup internet connection? 

Two bills do not automatically mean redundancy 

We sometimes come across businesses that have two internet services and reasonably assume that gives them redundancy. It might. The important part is how the two services reach the building and what they depend on upstream. 

If both connections use common infrastructure, the same building entrance or the same underlying carrier network, one incident can potentially affect both. 

So when we look at a backup connection, we are less interested in the fact that there are two invoices and more interested in what the two connections have in common. 

The backup does not always need to be as fast 

A company with a 1 Gbps primary circuit does not necessarily need another 1 Gbps circuit sitting idle. 

The first step is figuring out what has to keep working. For one business that might be the phone system, Microsoft 365 and a handful of cloud applications. For another, losing connectivity could stop a warehouse or prevent customers from placing orders. 

Once you know what has to survive an outage, you can size the backup around that requirement. 

A different technology can be useful 

One reason we use fixed wireless and LTE for some backup applications is that they can provide a different way into the building from a wired primary service. 

That does not make wireless automatically better than fibre, and it does not make every wireless connection independent. It simply gives us another option when we are trying to avoid a common point of failure. 

The connection is only part of the plan 

There are a couple of practical items that are easy to miss. The firewall has to know what to do when the primary circuit fails. If automatic failover is required, it needs to be configured and tested. The network equipment also needs power. 

And test it. We would rather find out during a planned test that something does not fail over properly than find out during an actual outage. 

How much redundancy is enough? 

There is no single answer. A small office that can work from mobile phones for an hour has a different risk profile from a company whose phones, orders and customer systems all stop when the internet stops. 

At iTeraTEL, we can look at the existing connection and the available alternatives before recommending a backup. Sometimes the right answer is quite simple. The important part is making sure the second connection actually addresses the problem you are trying to solve.