Production mail had stopped because the reverse DNS was wrong.
The forward records existed. The servers had addresses. What they did not have were the PTR records that let another mail system look up an IP address and confirm the name behind it.
Without that, messages were being rejected or delayed. The client's mail operation had reached a standstill.
I opened an urgent ticket with the colocation provider and attached the records that needed to be created. The list was long. Customer domains, mail hosts, application hosts, name servers, monitoring, load balancing, and internal services all sat across a small set of public addresses.
The public version of this story does not need that domain list. The useful fact is how a tiny DNS record could stop a much larger system.
Email trust is assembled from several checks. A server can accept a connection and still look suspicious. Forward DNS, reverse DNS, sender policy, reputation, and the behavior of the sending system all matter. Miss one layer and the message may never get through.
I sent the request several times, followed the provider ticket, and narrowed the immediate need to the mail host. Another person reported that turning on SSL in Outlook was helping some users, but that was a client-side workaround around a server-side problem.
The actual fix was not glamorous. Create the correct PTR record. Wait for the change. Test delivery. Make sure the mail starts moving again.
The subject line was exactly as calm as the outage allowed.
PTR records needed ASAP.