It strikes me that I haven't seen a general answer to the original
question - how to set up PTR records when one is serving more than one
domain under the same IP address.
This is of particular interest to me in that I currently do this as
well. What I'm doing now, seems to be working, but it's a matter of
accident, not design (small cluster, originally set up to support
company email and web servers for a few consulting clients, now also
hosting a variety of email lists -- the web servers all have their own
IP addresses, but the email domains share a common postfix installation
-- the postfix configuration and dns records have just been adjusted
over time). It's all working, nothing is getting blocked, but I'm not
sure why.
The original poster's question caught my attention - the RFCs suggest
that there should be only one PTR record per IP address -- which begs
the question of what do when one is serving multiple domains behind that
IP (be they virtual web servers or mail servers). And I can't seem to
find any established best practices (in RFC form or less formally) -
just a lot of anecdotal stories.
One thing that I've gathered is that how various programs - notably SMTP
servers and anti-spam packages - make use of PTR records, and how they
behave in the their absence, or in the case of mismatches, is idiosyncratic.
Which leads to several obvious questions:
- how does postfix use PTR records (e.g., which header lines are
matched, at what points in the processing chain, ...)?
- how does it react to the absence of a PTR record?
- how does it react to mismatches (and in which headers)?
- how much of this is configurable?
Yes, a lot of this is buried in the documentation - and I'm going off to
look - but the real question is: are there any lessons learned and/or
best practices to be applied to the general case of serving multiple
domains from the same IP address?
Inquiring minds want to know!
Miles Fidelman
--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra
