Lesson 20 · MX records - Build a DNS Resolver

MX records

An MX record names a mail server plus a preference number that orders competing servers. Today you parse its two-part RDATA - a 16-bit preference followed by a domain name.

The goal

Parse an MX record's preference and exchange name from its RDATA.

Start here - the target

TO DO

Scenario: An MX record is a preference then a name

Givenan MX record whose RDATA is 00 0A 04 6d 61 69 6c 07 65 78 61 6d 70 6c 65 03 63 6f 6d 00

Whenthe RDATA is parsed

Thenthe preference is 10 and the exchange is "mail.example.com"

Andthe preference occupies the first two bytes and the name fills the rest

Background

An MX record points mail for a domain at a mail server, and it has two parts: a 16-bit preference and an exchange name (the server). The preference orders alternatives - lower numbers are tried first - so a domain can list several mail servers with different preferences. On the wire the preference is the first two bytes of RDATA, and the exchange name follows immediately, encoded just like any other name and eligible for compression.

So parsing MX is: read a 16-bit value, then decode a name starting two bytes into the RDATA (against the full message, so pointers resolve). The RDATA here gives preference 10 and exchange mail.example.com. This preference-then-name shape is a common pattern - a couple of fixed fields followed by a name - and it is good practice for the multi-field record coming next.

Make it work

type MX struct {
  Preference uint16
  Exchange   string
}
// first two bytes are the preference, then a name (which may compress)
pref := uint16At(rdata, 0)
name, _ := decodeName(msg, rdataOff+2)

CheckpointDONE

You can parse an MX record. Commit and stop here.