High number of messages in Inbound or Outbound queue

Products

• Messaging Gateway

Issue/Introduction

Email delivery is being deferred by Symantec Messaging Gateway (SMG) with the following SMTP response: 421 4.3.1 queue is full

This issue occurs when there is a high volume of messages residing in either the Inbound or Outbound message queues. These messages are not assigned an Audit ID because they were never accepted by the scanner, though the deferred connection attempt is logged for the connecting IP address.

Cause

The "queue is full" state is triggered when applications or email clients send messages faster than the SMG scanner can process them. This causes the queue to reach its maximum capacity.

Contributing factors include:

• Traffic Direction: This can affect both Inbound (external mail) and Outbound (internal/application mail) queues.

• Message Size Processing: Increasing the accepted size limit of messages causes the scanner to spend more time on deep-content inspection for each item. This slowdown prevents the queue from clearing at a normal rate.

Resolution

To address the high number of messages in the queue and prevent future deferrals, follow these recommendations:

• Optimize Message Size Limits: Increasing the accepted message size can cause the scanner to take too much time processing individual messages. It is better to keep the default size limit at 10.4 MB to ensure optimal scanning performance and rapid queue throughput.

• Scale Scanning Resources: Ensure that traffic is distributed across multiple scanners. Binding too many high-volume applications to a single scanner is not recommended and is a primary cause of queue exhaustion.

• Throttle Application Dispatching: If you cannot deploy additional scanners, adjust how applications send mail. Instead of sending large bursts of data at once, stagger the delivery (e.g., sending a set amount of mail every minute) to allow the scanner to keep pace.

• Monitor Queue Status: Regularly monitor the "Message Queues" in the Control Center to identify if the bottleneck is specifically on the Inbound or Outbound side to better target your troubleshooting.