With the recent Nord Stream explosions and war in Ukraine, sabotage-type attacks on infrastructure are occurring more frequently and causing more damage. Most recently, on October 19, an Internet cable in the south of France was severed, impacting subsea cable connectivity to Europe, Asia, and the United States, and causing increased website response latency.
Undersea internet cables have become a national security risk and can be the cause of significant outages for enterprises that are not architected properly. Today, technology is being created with the sole purpose of attacking and destroying subsea cables, so what we’ve seen thus far is only the beginning of what’s to come.
Internet connectivity for both businesses and individuals worldwide relies heavily on undersea cables. When the Asia-Africa-Europe-1 (AAE-1) Internet cable was cut on June 7, for example, millions of people lost Internet connection. The country of Ethiopia lost 90% of its connectivity, while Somalia lost 85% of its connectivity. Tech giants in cloud computing like Google, Amazon, and Microsoft all had service interruptions as well. Given this reality, it’s imperative that network architects consider these types of failures and how they can be addressed. Businesses that are not preparing for the inevitable uptick in service interruptions due to subsea cable attacks are doing a disservice to their employees and, most importantly, their customers.
Part of the problem is that the Internet is generally resilient, so businesses – depending on where they’re located and who they most frequently interact with – haven’t had to worry about these types of issues on a large scale before. These businesses have become accustomed to a scenario where traffic is automatically rerouted, often in the blink of an eye, if one cable stops working. However, with more strategic targeted attacks on major chokepoints, what was once a hiccup could be several hours without Internet connectivity. This can cost businesses thousands or even millions of dollars, and, in an always-on world, it can cause reputational damage as well.
Fortunately, enterprises – regardless of how they’re currently architected – can better protect themselves before the next attack occurs, because there are solutions out there that have been designed to handle these types of challenges.
Enterprises should look at their current service level agreements with their network or telecommunication provider. What may have sufficed previously may no longer be adequate for the current environment. Re-evaluate key categories such as packet loss, latency, jitter, and fail-over. If the numbers don’t stack up and there are better alternatives at a lower rate, now is the time to renegotiate current agreements or consider one of the alternatives. There have been drastic improvements and innovations in network architecture over the years that can truly make a difference when the next attack takes place. Continuing to use what you already have because "that's the way things have always been done" is not a good enough reason to maintain your current approach.
For example, a network with a meshed, private global backbone with automatic rerouting can ensure that if something catastrophic does happen, end users would not experience service degradation related to latency, packet loss, or jitter. This means that mission-critical applications can remain operational without a hitch. This is possible because of Layer 1 links that exist between continents, where customer traffic can be prioritized over internet traffic.
Additionally, hyperscale points of presence (PoPs) are constantly analyzing link status and adjusting as needed. It’s important to note, though, that this is not always automatic, and will depend on the service provider. In some cases, even if a given PoP has multiple resilient links, it must also be able to handle the customer traffic after it automatically reroutes around failures. If a major chokepoint is destroyed, this can cause a bandwidth issue that can delay the rerouting process in certain instances.
Quickly rerouting traffic is more difficult with more traditional solutions like MPLS, which wasn’t designed with the cloud in mind, because paths are predetermined and can be less dynamic. There are times when the predictability of a traditional model can be helpful, but when dealing with these modern, calculated attacks that are aiming to cause the most damage, they fall short of what businesses require.
While the recent uptick in sabotage-type attacks on subsea cables is vastly different from the pandemic-caused disruption, which happened quickly and had long-lasting effects, these attacks are still similar as they have a growing impact on how businesses connect with customers and the world around them. Relying on the systems of yesteryear to get the job done is a recipe for disaster because those systems were built on certain assumptions that didn’t account for what we know now. Will there need to be multiple large-scale attacks before most enterprises seriously rethink their current approaches, or will the recent incidents be enough to spur them into action? In the game of cat and mouse, there are those who are proactive and strategic, there are those who are reactive to current circumstances, and there are those who continue to do the same thing over and over, no matter the consequences. Where does your business fall?
