To the sysadmin population of the Fediverse: do people have any numbers on how long common mail senders will retry sending mail if your MX is unreachable? Once upon a time people retried for many days, but my impression is that quite a few places now stop trying and bounce the email after quite short intervals, like a day.

(Boosts and practical experiences welcome, like "my MX was down for three days and I still got all that email sent from GMail".)

@cks 5 days should be still pretty common and the defaults on postfix and sendmail.

They'll send a notification mail every so often that it's undeliverable but it'll remain in the queue for up until that configurable limit to retry.

@matt It looks like the upstream Exim default is four days, and the Debian and Ubuntu configurations copy that in their split Exim setup (four days is honestly faster than I expected).

Mostly I'm wondering about the collection of 800 pound gorillas of GMail, Outlook, Amazon SES, etc etc, who might well have significant volume reasons to bounce email faster. (If spammers time out faster, well, that's a feature.)

@cks i have recently seen office365 (corp enterprise) retry for at least 48 hrs.

@cks

Before the days of preprocessors (i.e., Spam Assassin, Barracuda, etc.), notifications were sent to the catch-all admin account once every 24 hours, then 5 cycles(?) of notifications triggered it being marked as a 'dead letter', and dumped into a to-be-deleted later holding directory. I think those were generally held for 30 days. Server admins could retrieve and reroute them manually.

After preprocessors, things changed and all bets were off on expecting default timings.

@[email protected] my mail server give up on 4 days (exim default configuration)

@cks Will vary depending on temporary vs. permanent failure, but ppsw.cam.ac.uk would keep trying for two weeks, from memory.

(Permanent failures include, "Cannot resolve domain", which can interact excitingly with DNSSEC. So depending on the failure mode, the email might just immediately get returned regardless of retry policies. Also, not all senders send via a proper MTA. :/ Can be particularly problematic with website senders.)