Internet Outages On the Rise
Why are internet outages becoming so frequent today? More importantly, what can companies do to mitigate them? The pandemic certainly led to more people working from home and spending more time online than ever before. Could this have overwhelmed internet providers?
In a recent NS1 webinar, we dig deeper into the issue.
Keep reading to learn:
Why outages have grown more frequent - and where DNS comes into play
Key steps you can take at your company to make it less likely that you'll experience one - while also minimizing the impact if you do
The Role DNS Plays in Outages
More Devices Connected to the Internet = Increased Complexity
Modern life has become increasingly dependent on the internet. Today, more devices use the internet and DNS specifically to communicate than ever before. We're getting to the point where people may not be able to open their car door without a web server responding. So DNS has become very important to keep service up and live.
DNS seems simple and trivial until it breaks. When this happens, it can go spectacularly wrong. For example, in one recent incident employees could not get into their building to troubleshoot DNS issues... since the card reader at the building entrance was dependent upon DNS to work."
Beyond the number of devices now dependent on DNS, networks themselves have grown incredibly complicated, with layers of APIs. When one thing goes wrong, the result can be a cascading chain of errors. Sometimes the original cause is a server issue, other times it can be attributed to human error. Suddenly you've deleted all your domain and records, and people can't get in the building.
Watch the Full Replay of “It’s Always DNS”
Check out our recent Monthly INS1GHTS webinar, It's Always DNS: Why Outages Are Growing More Common, and What You Can Do About It here.
Then there are DDoS attacks. Imagine 20,000 people going to a store, all needing to go to the bathroom simultaneously. Regular customers can't even get in the door because everything is blocked. With DDoS attacks, it's important to calculate the cost of downtime to your company’s bottom line. How much will an outage cost? What happens when forced down by a bad actor? How much is it worth paying to avoid being in that situation? It's never a good time to have a DDoS attack.
So what is the best course of action in preventing downtime due to human error or DDoS attacks?
Steps You Can Take to Prevent Service Disruptions
Automating routine tasks and maintenance is a great way to minimize the likelihood of human error causing an outage. Additionally, implementing automated traffic steering policies - in combination with real-time monitoring - enables you to proactively steer traffic away from unhealthy endpoints. If, for example, one of your CDNs is experiencing an outage, this can help you minimize the likelihood of service disruptions or worsening response times for your end-users.
It's valuable to note that you can use intelligent DNS traffic routing for internal resources as well as your external network. There can be an entire ecosystem of DNS within a network managing micro-services and various parts of the platform. In fact, many enterprise environments have grown so large and complex, managing them is almost equivalent to managing external network traffic.
Having dynamic distribution of infrastructure is now key. For example, it's increasingly common for 10 servers to be on a rotation. If three servers go bad, they can be pulled out of the rotation course, wiped clean, reinstalled with the OS, brought back up, and reinserted into the loop. Having the right redundancy allows systems to bring forth cutting-edge features capable of handling the newest, most interesting, and complex issues.
Enterprise companies, regardless of size, shouldn't wait until things break to focus on improving resilience. In general, assume there will be a failure at some point and try to leverage various technologies with DNS routing to create self-healing, self-scaling systems that are resilient. If one server goes down, there should be a system to prevent everything from collapsing.
Having redundancy in place is crucial to keeping a site up during outages and DDoS attacks. Brainstorming in preparation of what can go wrong and finding out what redundancy is already in place is important. If the DNS isn't working and nobody can remember the IP address for the site to function, that's an issue that needs to be prioritized to succeed today.