It seems that when someone sends a large number of SMTP messages, all the server’s delivery threads get tied up delivering to SMTP addresses (which is slow, especially when Retries occur) - no threads are available to route NRPC messages. (19 threads are active)
Mail quickly backs up to other users in the domain. I mean, hundreds and hundreds.
Tell Router Sh Qu seems to indicate that some SMTP servers are being retried while others don’t appear to ever have been tried (including NRPC), instead of rolling thru all servers before starting over.
Couple questions:
-
Is there a way to force all SMTP mail to run at a lower priority than NRPC mail?
-
Anyone else seeing this? It SEEMS that this never happened at 6.5.5 - and just started with 8.5, but of course I can’t prove it…