Originally posted on Forbes
When it comes to technology, you can’t stick to tried-and-true methods forever. The internet and everything dependent on it are constantly evolving. What may have worked five years ago could likely hurt you now.
Remember when there were buttons on phones? Dedicated MP3 players? These technologies that may have been revolutionary a few years ago are now obsolete.
The same goes for the internet and the services we use to manage our businesses on it. DNS is a 30-year-old protocol that still uses the same IP to domain resolution three decades later, but nearly everything else about how we manage it has changed.
Just a few years ago, routing was limited to the health of your IP address (i.e., is it up or down?). Now, thanks to cloud technology, we’ve discovered that you can route clients based on where they are or what networks they are using all in real time. That means the client making the DNS query will be answered with a unique response every time.
If there are so many exciting new things going on in the DNS space, why aren’t more CTOs and CIOs talking about them? Simply put, we’ve seen DNS be too often discussed only when something goes wrong.
DNS Is An Afterthought
This brings us to our first outdated practice: thinking of DNS as an afterthought. For most businesses, it takes a disaster for decision makers to revisit how they are managing their DNS and seek out a new solution.
Unfortunately, many decision-makers don’t think any more of DNS beyond being a small part of the stack. While this is true, DNS lies at a critical ingress point that holds up the rest of your stack. Therefore, DNS hosting needs to be prioritized and delegated to a trusted provider.
The problem is, businesses that have other specialized cloud hosting needs (like content delivery networks or e-commerce enterprises) will get sucked into using their provider’s DNS services. This has some advantages, such as integrations with the other services and only needing one account. However, putting all your eggs in one bundled basket could cost thousands or even millions of dollars of lost revenue –and your brand reputation — if that provider is downed.
We think the best way to ensure availability and optimal performance for your domains is to spread your hosting needs across multiple providers that specialize in that given service for increased redundancy.
Who Needs Redundancy?
The second bad practice, which is still being used by the majority of top domains, is single homing DNS management. Companies that relied on a single DNS provider suffered massive DNS outages after the famed worldwide outagelast October. A major DNS provider was downed by a DDoS attack, which in turn knocked thousands of large domains offline. Only a select few were able to withstand the outage because they were using more than one DNS service.
This is a relatively simple configuration that automatically distributes your DNS traffic across multiple providers’ servers. In the event that one provider is unavailable, the secondary provider answers all of the traffic.
We’ve discovered a few different ways to do this: You can have two primary providers, one primary and one secondary, or, as many companies are now opting to do, choose services that will automatically integrate with additional providers.
If it’s so easy, then why are over half of the top 100 domains still using only one provider?
Some organizations like to think they are immune. Whether it’s size, clout or something else entirely, it doesn’t matter: You still need redundant DNS services or you will go down. Since last May, three major DNS providers (Dyn, NS1 and GoDaddy) have suffered downtime and saw vast swaths of domains go down with them.
Take The Easy Way
Last, but definitely not least, CTOs and CIOs are not doing enough research when looking for service providers. It doesn’t matter if it’s DNS, CDN, web hosting or an email client. You need to conduct your own research and testing. Create test domains and test as many of the services as you can. You should feel comfortable when you move your business over to a provider’s servers, not worried.
If you don’t have time to do your own testing, then do your due diligence and find other trusted sources that have done the research for you.
Back to our original question: Why are CTOs still using old DNS practices? Simply put, it’s easier. Finding the extra time to revamp your DNS can be challenging. On top of that, you have to do double the research of finding not one but two good providers and then test everything to make sure migration is smooth. But once you’ve done this, you can help save your business time and money in the long run.
Also published on Medium.