Transactional email silently stops going out and the mail queue keeps growing. In Postfix that almost always means deliveries are deferring — the server accepted the mail but can't hand it to the next hop, so it keeps retrying on a backoff while the queue piles up.
Read the queue and the log
mailq | tail # or: postqueue -p (how deep, and to where) tail -f /var/log/mail.log # status=deferred (connect to relay timed out) = relay is the problem
The status= text is the whole story: deferred with a connection timeout points at the relayhost; bounced with a 5xx points at auth or recipient problems.
Fix the relay, then flush
# /etc/postfix/main.cf relayhost = [smtp.relay.internal]:587 postfix reload # pick up the config postqueue -f # flush: retry every deferred message now
Don't just postqueue -f on repeat — flushing before you fix the relay only hammers a dead host and resets the backoff. Fix the cause first, then flush once.
Practise it
A deferred mail backlog is one of the simulator's mail-tier incidents. Work it in a full incident response shift.