Len, thank you for your deep answer. I learned a lot. 

regards

        -Luis Arango

> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:IMail_Forum-
> [EMAIL PROTECTED] On Behalf Of Len Conrad
> Sent: Martes, 24 de Mayo de 2005 03:30 p.m.
> To: [email protected]
> Subject: RE: [IMail Forum] How Imail handles delivery when the MX record
> has multiple IPs
> 
> 
> >Sandy: I did. If tests are correct, it seems like Imail doesn't lookup
> for
> >the next IP in the MX record when various IPs are registered in the MX
> >Record. Instead it goes to the next MX record.
> 
> The MX algorithm says all MX records at a given preference must be tried
> before moving to the next lower preference MX.
> 
> The MTA resolves an MX hostname to an A record, and tries it. The A record
> selected to be tired is usually only the first physical A record in the
> answer packet, if a record-set of A records in an an DNS answer packer.
> 
> here's the example of a domain having a record-set of MX records:
> 
> %dig aol.com mx
> 
> ; <<>> DiG 9.2.3 <<>> aol.com mx
> ;; global options:  printcmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 45549
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 4, AUTHORITY: 4, ADDITIONAL: 18
> 
> ;; QUESTION SECTION:
> ;aol.com.                       IN      MX
> 
> ;; ANSWER SECTION:
> aol.com.                3534    IN      MX      15 mailin-03.mx.aol.com.
> aol.com.                3534    IN      MX      15 mailin-04.mx.aol.com.
> aol.com.                3534    IN      MX      15 mailin-01.mx.aol.com.
> aol.com.                3534    IN      MX      15 mailin-02.mx.aol.com.
> 
> all MX records of equal value, so all MX hostnames are to be tried,
> usually
> in the physical order of the DNS answer packet above.
> 
> The MX algorithm doesn't address the case of a A record-set for an MX,
> such
> as this:
> 
> %dig  mailin-03.mx.aol.com.
> 
> ; <<>> DiG 9.2.3 <<>> mailin-03.mx.aol.com.
> ;; global options:  printcmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 18764
> ;; flags: qr rd ra; QUERY: 1, ANSWER: 4, AUTHORITY: 2, ADDITIONAL: 2
> 
> ;; QUESTION SECTION:
> ;mailin-03.mx.aol.com.          IN      A
> 
> ;; ANSWER SECTION:
> mailin-03.mx.aol.com.   288     IN      A       64.12.137.249
> mailin-03.mx.aol.com.   288     IN      A       64.12.138.120
> mailin-03.mx.aol.com.   288     IN      A       205.188.158.121
> mailin-03.mx.aol.com.   288     IN      A       64.12.137.152
> 
> The order of the A records is determined by the DNS answering the query,
> NOT by AOL's DNS. The order should be considered effectvely random and
> uncontrollable.
> 
> It's the MTA's implementation to decide to use only 64.12.137.249, before
> trying:
> 
> mailin-04.mx.aol.com.
> 
> ... or to try all the elements in the A record-set for mailin-
> 03.mx.aol.com.
> 
> 1. if the A record is not deliverable, then
> 
> 2. the MTA tries the "next physical" MX record (-04 above, its A record)
> at
> the same preference level
> 
> or
> 
> 3. tries the MX at lower preference.
> 
> 4. if no MX at lower preference, the MTA defers delivery for retry
> (re-starting the MX algorithm from the top at the next retry)
> 
> >I just want to confirm this is a behavior by design, a problem or just
> miss
> >configuration from my side -probably it may be fixed by changing
> something
> >in the registry-.
> 
> You're worrying about something that is buried in the implementation of
> the
> MTA's MX algorithm, which itself is extremely well-known and ancient.
> There
> is NO controversy.  There are certainly variations in MTA implementation
> of
> the MX algorithm, but at this point, all the well-established MTAs have
> "got it right" for all intensive porpoises.
> 
> I personally have never had to get involved in an MTA's setting at this
> level in 6+ years of SMTP admin, so I'm not sure you're barking up the
> right tree, or if there is a tree.
> 
> 
> Len
> 
> 
> _____________________________________________________________________
> http://IMGate.MEIway.com : free anti-spam gateway, runs on 1000's of sites
> 
> 
> 
> To Unsubscribe: http://www.ipswitch.com/support/mailing-lists.html
> List Archive: http://www.mail-archive.com/imail_forum%40list.ipswitch.com/
> Knowledge Base/FAQ: http://www.ipswitch.com/support/IMail/
> ______
> [Email scanned for viruses by Panda Consulting -www.pandacons.com-]
> [Email escaneado contra virus por Panda Consulting -www.pandacons.com-]


______
[Email scanned for viruses by Panda Consulting -www.pandacons.com-]
[Email escaneado contra virus por Panda Consulting -www.pandacons.com-]


To Unsubscribe: http://www.ipswitch.com/support/mailing-lists.html
List Archive: http://www.mail-archive.com/imail_forum%40list.ipswitch.com/
Knowledge Base/FAQ: http://www.ipswitch.com/support/IMail/

Reply via email to