Hi,

I did now some more scans for MTA-STS and I thought it might be
interesting for the list to learn the results.

A very effective way of finding hosts that support mta-sts is to scrape
the Certificate Transparency logs. (With the exception of hosts that
use wildcard certificates.)

This gave me 697 hosts with an mta-sts subdomain. Of those 416 served
something that looked like an mta-sts policy file, indicating that a
large number (281) are either in the process of deploying MTA-STS and
haven't finished yet or have wrongly implemented it, e.g. by using the
wrong filename/path.

I found a few syntax issues:
* The most worrying one is that 24 hosts use policies like
  "mx: .example.org" which was valid in older drafts. I say this is the
  most worrying, because it may actually lead to delivery failures.
  It'd be good to get them converted quickly before this creates
  hassle. However only 4 of them have "mode: enforce" (with "mode:
  testing" I'm not overly worried).
* 11 hosts use "mode: report", which is also from earlier drafts and
  should be "mode: testing". However a number of hosts have already been
  fixed before I made this scan.
* Some have trailing or leading spaces. It's not entirely clear to me
  if this is something to be considered an invalid policy.
* I also wonder about line endings. 263 use Unix line endings (LF
  only), 154 use Windows line endings (CRLF). The RFC reads like CRLF
  is correct (3.2), which would indicate a large number of bad policies.
  However the formal definition in the RFC also lists <LF>. I am not
  familiar with how to read the formal definition. (It also says
  something about spaces.)

I'd appreciate if someone could clarify whether leading/trailing spaces
and unix line breaks should be considered policies that should be
fixed or if this is okay.


Some stats about modes:
211 mode: enforce
194 mode: testing
11 mode: report (invalid)
no mode: none found.



-- 
Hanno Böck
https://hboeck.de/

mail/jabber: [email protected]
GPG: FE73757FA60E4E21B937579FA5880072BBB51E42

_______________________________________________
Uta mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/uta

Reply via email to