Mail queue started filling up on one of the Ironports after this alert.
The url_rep_client was restarted due to an error.
Looking at the logs I see this error.
grep “unable to” mail_logs -t
ACCEPT SG None match sbrs[none] SBRS unable to retrieve country unable to retrieve
DNS checked out ok. Found a TAC case that indicated a possible bug. Forced update on reputation engine.
Confirmed update by tailing the updater log file.
Forced mail delivery in the queue.
Please choose an option for scheduling immediate delivery.
1. By recipient domain
2. All messages
No change. Decided to reboot with 30 seconds to wait before forcing connections to close. I was a bit worried about all the stuck messages in the queue. After the reboot queue retained all the original messages and slowly started delivering them.
Sometimes all you need is a good old system reboot.