[114752] in North American Network Operators' Group
MX Record Theories
daemon@ATHENA.MIT.EDU (gb10hkzo-nanog@yahoo.co.uk)
Tue May 26 14:04:12 2009
Date: Tue, 26 May 2009 11:03:59 -0700 (PDT)
From: gb10hkzo-nanog@yahoo.co.uk
To: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Hello all,=0A=0AFirst, I hope this is not off-topic for NANOG, please be ge=
ntle with me as this is my first post.=0A=0AI=0Awould be most interested to=
hear NANOG theories on the variety of MX=0Arecord practices out there, nam=
ely, how come there seem to be so many=0Aways employed to achieve the same =
goal ? Do you have experience in=0Amore than one of these methods and whic=
h do you favour ?=0A=0ATo illustrate my question :=0A=0A(1)=0AIf you query =
the MX records for, Hotmail or AOL you will receive 4=0Aequal weight MX rec=
ords, each of the MX records having a round-robin=0Aset of IPs.=0Ae.g.=0Aho=
tmail.com. 2706 IN MX 5 mx4.hotmail.com.=0Ahotmail.com. =
2706 IN MX 5 mx1.hotmail.com.=0Ahotmail.com. 2706 IN=
MX 5 mx2.hotmail.com.=0Ahotmail.com. 2706 IN MX 5 mx=
3.hotmail.com.=0A-and-=0Amx3.hotmail.com. 1926 IN A 65.xxxxxxx=
=0Amx3.hotmail.com. 1926 IN A 65.xxxxxxx=0Amx3.hotmail.com. =
1926 IN A 65.xxxxxxx=0Aetc.etc.=0A=0A(2)=0AAlternatively, some peo=
ple, particularly the ones that use hosted=0Afiltering, tend to have one MX=
record, which as multiple round robin=0AIPs.=0Ae.g.=0Amicrosoft.com. =
780 IN MX 10 mail.global.frontbridge.com.=0A-and-=0Amail.global.=
frontbridge.com. 1728 IN A 65.xxxxxxx=0Amail.global.frontbridge.com. =
1728 IN A 207.xxxxxxx=0Aetc. etc.=0A=0A(3) And others simply have a m=
ore traditional setup using multiple MX records and only one IP per MX reco=
rd with no round robin=0Aapple.com. 931 IN MX 10 mail-in14.=
apple.com.=0Aapple.com. 931 IN MX 20 mail-in3.apple.com.=0A=
apple.com. 931 IN MX 20 eg-mail-in2.apple.com.=0Aetc.etc.=
=0A=0A=0ASo=0Awhat's the big deal ? Please note I'm not asking which is "b=
etter" ...=0AI am just curious and interested to hear your professional opi=
nions and=0Aexperiences.=0A=0APersonally, I favour the simple option 3, mul=
tiple MX records.=0A=0AThanks y'all.=0A=0A=0A
