| Commit message (Collapse) | Author | Age | Files |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This solves an issue where an attacker would strip the STARTTLS keyword
from the EHLO response, thereby preventing connection upgrade; or spoof
DNS responses to route outgoing messages to an attacker-controlled
SMTPd, thereby allowing message MiTM'ing. With key material pinning in
place, smtp(8postfix) immediately aborts the connection (before the MAIL
command) and places the message into the deferred queue instead:
postfix-out/smtp[NNN]: … dsn=4.7.5, status=undeliverable (Server certificate not verified)
This applies to the smarthost as well as for verification probes on the
Mail Submission Agent. Placing message into the deferred queue might
yield denial of service, but we argue that it's better than a privacy
leak.
This only covers *internal messages* (from Fripost to Fripost) though:
only messages with ‘fripost.org’ (or a subdomain of such) as recipient
domain. Other domains, even those using mx[12].fripost.org as MX, are
not covered. A scalable solution for arbitrary domains would involve
either DANE and TLSA records, or MTA-STS [RFC8461]. Regardless, there
is some merit in hardcoding our internal policy (when the client and
server are both under our control) in the configuration. It for
instance enables us to harden TLS ciphers and protocols, and makes the
verification logic independent of DNS.
|
| |
|
|
|
|
| |
Use admin@fripost.org instead. We were sending out (to the admin team)
system messages with non-existing or invalid envelope sender addresses,
such as <logcheck@antilop.fripost.org> or <root@mistral.fripost.org>.
|
| | |
|
| | |
|
| |
|
|
| |
Cf. lmdb_table(5).
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
These addresses need to be accepted on the MX:es, as recipients
sometimes phone back during the SMTP session to check whether the sender
exists.
Since a time-dependent suffix is added to the local part (cf.
http://www.postfix.org/postconf.5.html#address_verify_sender_ttl) it's
not enough to drop incoming mails to ‘double-bounce@fripost.org’, and
it's impractical to do the same for /^double-bounce.*@fripost\.org$/.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
In order to avoid ‘double-bounce@’ ending up on spammer mailing lists.
See http://www.postfix.org/ADDRESS_VERIFICATION_README.html .
|
| |
|
|
|
|
|
|
| |
Following Viktor Dukhovni's 2015-08-06 recommendation
http://article.gmane.org/gmane.mail.postfix.user/251935
(We're using stronger ciphers and protocols in our own infrastructure.)
|
| |
|
|
|
|
| |
Following Viktor Dukhovni's 2015-08-06 recommendation for Postfix >= 2.11
http://article.gmane.org/gmane.mail.postfix.user/251935
|
| |
|
|
|
|
|
|
|
|
| |
Ideally we we should also increase the Diffie-Hellman group size from
2048-bit to 3072-bit, as per ENISA 2014 report.
https://www.enisa.europa.eu/publications/algorithms-key-size-and-parameters-report-2014
But we postpone that for now until we are reasonably certain that older
client won't be left out.
|
| | |
|
| |
|
|
|
| |
Put all relay restrictions under smtpd_relay_restrictions and leave
smtpd_recipient_restrictions empty, since we don't do DNSBL.
|
| | |
|
| |
|
|
|
|
|
| |
We don't want to bounce messages for which the recipient(s)' MTA replies
451 due to some greylisting in place. We would like to accept 451
alone, but unfortunately it's not possible to bounce unverified
recipients due to DNS or networking errors.
|
| |
|
|
|
|
|
| |
This is specially useful for mailing lists and the webmail, since it
prevents our outgoing gateway from accepting mails known to be bouncing.
However the downside is that it adds a delay of up to 6s after the
RCPT TO command.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
We already removed it from the MX:es (see 32e605d4); we need to remove
it from the MDA and outgoing SMTP as well, otherwise mails could bounce
or get stuck in the middle (the're rejected with 450: deferred by
default).
However we can keep the restriction on the entry points (MSA and
webmail).
|
| | |
|
| |
|
|
| |
For DKIM signing and virus checking.
|
| |
|
|
|
| |
For some reason giraff doesn't like IPSec. App-level TLS sessions are
less efficient, but thanks to ansible it still scales well.
|
| |
|
