Hi All,
We're running 4 exchange servers. All are on Server 2016 with Exchange 2013 installed.
The servers are as follows:
EX01 - Ver. 2013 Set in a VIP to be the main server/IP for outlook in DNS)
EX02 - Ver. 2013 (Will be decommissioned as it is not going to be required in the future)
EX03/EX04 - Both Ver. 2013 (Configured as a DAG pair with a separate file server as witness, 03 holds all primary DB's with 04 having the secondary copies)
Over the past few days I have noticed that the Shadow Redundancy queue in EX03 is ballooning fast in size with emails stuck not sending to what I assume is EX04 from the name of tab in the Queue Viewer. The queue show all the emails with a status of "Ready" and there is no "Last Error" in the individual properties of each.
Consequentially all of our servers have suddenly started showing RAM usage of near 100% constantly and when viewing the "SMTP Relay to Database Availability Group" queue from EX01 to the DAG is showing stuck emails with a system memory limit reached error.
I have no restarted the transport service, that will be done during a change window the following day.
I guess my questions to everyone here is:
1) Is there any correlation between RAM/Memory capacity and the Shadow Redundancy Queue blowing out?
2) Is there a way to pinpoint what is failing within the DAG/Routing to cause the shadow queue not to clear? I have confirmed with users that they are receiving the emails that are stuck in the queue so the ACK sent back does not reach the server it seems?
Appreciate your time for reading this and apologies if I've missed anything, happy to answer any questions you've got.