The Need for Speed: Highlights from IBM and Catchpoint’s Global DNS Performance Study

Despite DNS being the backbone of Internet connectivity, reliable metrics for benchmarking DNS performance are surprisingly scarce. This gap often leaves IT teams navigating in the dark, unable to effectively gauge how their DNS configurations stack up against industry standards. To address this pressing need, Catchpoint worked with IBM NS1 Connect to provide a clear, data-driven picture of DNS performance. This collaboration was aimed at dissecting DNS performance across the globe, using data from some of the most visited websites to reveal how DNS performs “in the wild.”  

The Study: Methodology and approach

The study encompassed 2,271 of the most visited websites across the world. This diverse pool was tested using our industry-leading Global Observability Network and robust DNS monitoring capabilities.  

Here’s what we found:

#1 - Benchmark for DNS performance

The average DNS response time across the 2,271 websites studied was 263 milliseconds. This global average provides a benchmark for understanding typical DNS performance across diverse Internet environments. However, there is a lot of regional variation in that global average.

#2 - DNS performance varies significantly by region

The study found significant regional differences in DNS performance, with different continents showing varying results.

• North America and Europe showed faster response times, mainly due to the higher density of DNS servers.
• Asia, South America, and Oceania experienced slower DNS responses, influenced by greater distances and fewer DNS servers nearby.

These variations highlight the impact of geography and infrastructure on DNS performance. Regions with denser server networks and shorter distances between users and servers tend to perform better. IT teams should consider these factors when evaluating DNS performance and develop region-specific strategies.

#3 - Managed vs. self-hosted DNS

One of the starkest contrasts we uncovered was between self-hosted DNS setups and managed DNS services:

• Self-hosted DNS was 35% slower than the average global response time, with an average difference of 141 milliseconds.
• Compared to managed DNS providers, self-hosted DNS lagged 41% behind the global average and 60% behind IBM NS1 Connect specifically.

This data suggests that the control offered by self-hosting doesn’t compensate for the performance drawbacks, making managed DNS a more practical option for most businesses. The report states, “Many companies decide to self-host their DNS to control everything about their connections, but our numbers clearly show that the investment isn’t paying off when it comes to performance.” Therefore, opting for a managed DNS solution is the best option rather than going at it alone.

#4 - Not all managed DNS services are created equal

The study found significant performance differences among DNS providers, emphasizing the need for businesses to evaluate their vendors carefully:

• IBM NS1 Connect emerged as the leader, with response times 39% faster than the global average.
• Vercara UltraDNS followed closely, being 35% faster.
• Cloudflare DNS came in third, outperforming the global average by 16%.

On the other hand, AWS Route 53 and Akamai Edge DNS fell below the global average, with AWS being 9% slower and Akamai 24% slower, highlighting the variability even within premium services.

#5 - Traffic volume misconceptions

One might expect that the busiest websites with the most sophisticated networks would consistently show superior DNS performance. Surprisingly, the distribution of DNS response times was more uniform than anticipated:

• High-traffic websites did not always perform better than their less-visited counterparts.
• Some top-tier websites lagged behind, while many lower-tier sites performed exceptionally well.

Only 23 companies reported DNS response times slower than 1,000 milliseconds, most of which used self-hosted DNS solutions.

So, what do all these findings mean for your business when formulating a DNS strategy? How did we benchmark DNS performance? How do our tests compare to other DNS performance measurements in use today? Download the full report to learn more (no registration required).