The Story Behind Route 53's Name and DNS Trivia - Why Port 53, and Why UDP

Starting with how Route 53 got its name from port number 53 and America's Route 66, this article explores why DNS uses UDP port 53, and what happens behind the scenes during DNS resolution.

The Double Meaning Hidden in Route 53's Name

The "53" in Amazon Route 53 comes from TCP/UDP port number 53, the port used by DNS. DNS requests and responses use UDP port 53 by default. The "Route" part represents DNS's function of "routing" domain names to IP addresses, while also paying homage to America's famous highway, Route 66. Route 66 was the historic arterial road connecting Chicago to Los Angeles, known as "America's Main Street." Route 53 similarly carries the meaning of being the "main artery of the internet," guiding traffic to the right destination. Several AWS service names share this kind of dual structure, combining technical and cultural references. Incidentally, the real US Route 53 is a short state highway in Wisconsin, far less famous than Route 66.

Why DNS Uses UDP Port 53

The reason DNS uses UDP dates back to the network environment of 1983, when DNS was designed. Internet bandwidth was extremely limited at the time, and the overhead of TCP's three-way handshake (SYN, SYN-ACK, ACK) was a non-trivial cost. DNS queries and responses are typically small, under 512 bytes, and complete in a single round trip. TCP requires about 9 packets total (3 for connection setup, 2 for data exchange, and 4 for teardown), while UDP needs only 2 packets (query + response). The assignment of port number 53 is tied to the history of IANA (Internet Assigned Numbers Authority) port allocations. When DNS was standardized in 1983 through RFC 882 and RFC 883, port 53 was still available and was assigned accordingly. Port numbers 1-1023 are called "well-known ports" and are reserved for major protocols. Just as HTTP uses 80, HTTPS uses 443, and SSH uses 22, DNS uses 53. However, when a DNS response exceeds 512 bytes (such as DNSSEC responses), DNS falls back to TCP.

What Happens After You Type a URL in Your Browser

Between typing example.com in your browser and seeing the page, DNS resolution goes through multiple stages. First, the browser checks its own DNS cache. If there is no cache hit, it checks the OS DNS cache (mDNSResponder on macOS, systemd-resolved on Linux). If the OS cache also misses, a query is sent to the configured DNS resolver (usually your ISP's DNS server or a public DNS like 8.8.8.8). The DNS resolver checks its own cache, and if it misses, begins recursive name resolution. First, it asks a root DNS server (13 clusters worldwide) "Where is the authoritative DNS server for .com?" Next, it asks the .com authoritative DNS server "Where is the authoritative DNS server for example.com?" Finally, it asks the example.com authoritative DNS server (ns-xxx.awsdns-xx.com in the case of Route 53) "What is the IP address of example.com?" These three stages of queries represent the full picture of DNS resolution when there is no cache. Each stage's response includes a TTL (Time to Live), and the cache remains valid for the duration of the TTL.

Route 53 Alias Records - An AWS-Original Invention

Route 53's alias record is an AWS-proprietary feature that does not exist in the DNS standard. Standard DNS does not allow CNAME records at the domain apex (Zone Apex, e.g., example.com). RFC 1034 specifies that CNAME records must not coexist with other record types at the Zone Apex. However, CloudFront distributions and ELB endpoints are provided via dynamic DNS names (like d1234.cloudfront.net), so pointing the Zone Apex to these services requires a CNAME. Route 53's alias record was invented to solve this problem. An alias record resolves the DNS query internally within Route 53 and returns the final IP address as an A record. From the client's perspective, it is indistinguishable from a regular A record. Another advantage of alias records is that alias queries from Route 53 to AWS resources (CloudFront, ELB, S3 website hosting, etc.) are free. Standard DNS queries cost $0.40 per million, but alias queries incur no charge. Azure DNS and Google Cloud DNS do not have an equivalent to alias records, and each provider addresses the Zone Apex CNAME problem with different approaches.

Why DNS Outages Can Cascade Across the Entire Internet

DNS is the most fundamental layer of the internet, and when DNS goes down, virtually all internet services become unreachable. During the massive DDoS attack on Dyn DNS in 2016, services including Twitter, Netflix, GitHub, and Spotify were inaccessible for several hours. These services themselves were running normally, but users could not resolve the services' IP addresses because DNS was not responding. Route 53 has several countermeasures against this type of failure. First, Route 53's authoritative DNS servers operate on an Anycast network distributed worldwide, making it structurally difficult for an attack on a specific location to cascade to the entire system. Second, Route 53 is one of the few AWS services that offers a 100% availability SLA. This SLA guarantees that DNS queries receive correct responses. Third, Route 53's health check feature enables DNS failover, automatically redirecting traffic to healthy servers when backend servers experience failures. To systematically learn about DNS mechanics and network design, specialized books (Amazon) are a great reference.