I’m finding it difficult to give this post a proper heading but hopefully the tags in the post would have led you to the solution of this problem I experienced.
We have a specific Send Connector in Exchange 2016 that sends mail to one of our other companies in a different forest. I noticed that only on one server the queue for mail destined to this domain was queueing.
The error in the queue destined to the other domain was stating that it failed to connect. I tested with telnet to the VIP IP and could not connect to the VIP on port 25.
After restarting the Transport service and rebooting the server since only this server was experiencing the issue, I found that mail was still stuck in the queue to this VIP.
We checked the F5 and eventually found that I was unable to ping the gateway from my Exchange server but could ping it from my other Exchange nodes.
Changing the gateway to another gateway solved the issue but changing back to the original gateway caused the queue build up again. Scratching our heads that the gateway works for the other nodes.
The solution at the end was to clear the arp cache on the server where the mail was queueing up. From the command prompt I ran arp -a to get the cached entry for the gateway and then ran arp -d to delete the arp cache.
The issue was resolved after we cleared the arp cache from this Exchange node.