Originally posted in Forbes
If you’ve ever had to change which hosting provider you use or add a CDN (content delivery network), then you most likely have used a DNS management service before. It’s a relatively straightforward process: add your domain(s) to the new name servers, notify your registrar, create records, define rules and forget about it until something goes awry.
This was how most administrators saw DNS management — that is, up until a few years ago, when DNS monitoring hit the market. DNS monitoring allows you to see your query traffic in real time. These tools are heralding a new breed of management that shifts from reactive to proactive.
This new technology has changed DNS management from a one-and-done utility to an invaluable tool used frequently to optimize network performance. Admins are able to use query analytics to anticipate and solve issues before their end users are affected.
Before query analytics reached the mainstream market, old-school query monitoring was almost exclusively used by network ops to identify DNS-level attacks. This meant hours of sifting through backlogs.
Modern query monitoring services do all the hard work for you and present your query logs in interactive charts and tables. Some tools even allow you to filter queries by location, record type and IP protocol. Query analytics can give you a better sense of query density based on end users’ locations. They have also reportedly helped network engineers identify and anticipate DNS-based attacks like DDoS (distributed denial of service) attacks.
Real user monitoring (RUM), another type of monitoring service, turns queries into actual users. You can compare query resolution times based on end-user attributes like connection type (Wi-Fi or cellular), browser, location and more. RUM can also be used to troubleshoot connectivity issues and improve resolution times on a global scale.
You can then take the analytics and apply what you’ve learned to your DNS configurations. This way, you create unique experiences for your users specific to the way they access your domains.
The best way to visualize this change is to compare it to how we navigate in traffic. Before we had Google Maps or GPS systems installed in our cars, all we had were maps. If you were driving down a major highway and traffic suddenly came to a halt for no reason, there was no way for you to avoid it and no way to know you should’ve taken an alternate route.
This is your run-of-the-mill DNS management. If you’ve ever resorted to using your registrar’s DNS hosting, you know what we’re talking about.
Let’s jump forward a few years to the age of Google Maps and Waze. If you have ever lived in any metropolitan area, you can’t live without them. That’s because GPS apps gather real-time traffic conditions and dynamically update your directions to ensure you are on the fastest route.
Managed DNS services are following suit and have begun to offer the same functionality but for query routing.
These services harness the data from dozens of monitoring nodes to show real-time traffic conditions. This information is then used to inform routing decisions so end users are always answered quickly.
Let’s say you have a dense user base on the East Coast. Your DNS provider likely has multiple points of presence (data centers) across the region, which can easily handle the day-to-day traffic. But on Black Friday, your East Coast traffic comes to a crawl.
You can create rules that will automatically reroute some of your East Coast users to data centers in the Midwest. Under normal traffic conditions, this could add a few milliseconds to resolution times, but in this case, users would enjoy usual speeds and you would reduce congestion, thus improving resolution for East Coast users.
These are only a few of the applications of query monitoring to improve global performance. We recommend that you take a few minutes to audit your own organization’s online presence and look for bottlenecks in traffic, localized congestion or even unexplained latencies. Query analytics can help you not only identify but also troubleshoot these issues quickly and efficiently.