Re: IPv6 recap
Alessandro Vesely <[email protected]>
| Newsgroups | gmane.mail.imap.courier.general |
|---|---|
| Message-ID | <[email protected]> |
On Fri 05/Apr/2024 23:28:41 +0200 Sam Varshavchik wrote: > Alessandro Vesely writes: >>> >>>> this happened again, but this time caused a bounce: >>>> >>>> <<< 554 resimta-c2p-559421.sys.comcast.net >>>> resimta-c2p-559421.sys.comcast.net 2a02:29e1:500:6c00::4 Comcast requires >>>> that all mail servers must have a PTR record with a valid Reverse DNS >>>> entry. Currently your mail server does not fill that requirement. For more >>>> information, refer to: https://postmaster.comcast.net/smtp-error-codes.php#554 >>>> >>>> Is it possible to inhibit use of IPv6 a little bit harder? >>> >>> Can you block all IPv6 connections to port 25 on your firewall? >> >> Yes, I could. However, that way I'd lose the possibility to reach IPv6-only >> MXes (if there are any). >> >> If that depends on DNS queries timing, it should be enough to enlarge the >> timeout a little bit. Is that configurable? > > Nope, but it won't be difficult to make it configurable to some degree. > Handling of DNS query timeouts already has some logic to it, there's going to > be a retry if there's no response, and all nameservers from resolv.conf will be > tried, then if there's no response there will be additional rounds, with > progressively higher timeouts. This can be tweaked as long as the timeout is > the same for all DNS queries. It'll be harder to make the timeout different > depending on what's being queried. > > I don't know if this will help though. The baked-in logic makes three attempts > using an exponential backoff, timing out each attempt after 5, 25, then 125 > seconds. Per resolver. The last, 125 second timeout, is pretty generous. It happened again. Yesterday I sent 19 messages to Comcast, 18 were successfully delivered, one failed after trying to use IPv6. I'd guess Comcast's server answers after about 4.9 secs, so that when the first query randomly fails Courier tries the second address. I tried to look back at the rfc1035 code to recall how it works. It seems that if A and AAAA are queried simultaneously, the query fails unless a response is received for both. In that case, the choice between IPv4 and IPv6 would be deterministic, wouldn't it? Should I try a patch? Best Ale -- _______________________________________________ courier-users mailing list [email protected] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users