4 ms·

I now feel really really bad for fuck@off.com. I owe someone an apology...

• icebraining 15y ago
  You're fine, off.com doesn't have MX records.
  • holdenk 15y ago
    Except some people fall back to the A record (see RFC 5321).
    • seanp2k2 15y ago
      Can you please let me know where you're seeing this? I looked through http://tools.ietf.org/html/rfc5321 http://tools.ietf.org/html/rfc5321 for /\sa\s/i and /(fall(\s)?back)/i and didn't find anything related to this this behaviour. I ask because I'd like to have my answer at http://stackoverflow.com/questions/8221381/where-does-email-sent-to-example-com-go/8221408#8221408 http://stackoverflow.com/questions/8221381/where-does-email-... be as accurate as possible (plus I'm now curious about this). Thanks!
      • there 15y ago
        it's not a fallback, it's the default behavior. originally when you would send mail to a user at a host, the MTA would just connect to that host on port 25. the original SMTP RFC (821) pre-dates any DNS RFCs, so originally there wasn't even a concept of MX records and you just e-mailed someone at the server they had an account on. after DNS was in place, MX records came along in order to route mail destined for a host to a different server, or just supply a list of backup servers. now since most people just use email addresses containing only a domain, MX records are pretty much common place (since the A record of many domains resolves to the web server). now MTAs check for MX records before trying to connect directly to the host. to demonstrate: jcs@thalamus:~> host -t mx test.jcs.org test.jcs.org has no MX record jcs@thalamus:~> host test.jcs.org test.jcs.org has address 10.10.10.10 jcs@thalamus:~> echo test | mail test@test.jcs.org and shortly after, in postfix's mail log: Nov 21 22:59:18 thalamus postfix/smtp[23742]: connect to test.jcs.org[10.10.10.10]:25: Operation timed out