Number | 1874 |
Severity | Critical |
SNMP Name | ptDSCNodesRequestRejectedQueueFull |
Description | Within the last 5 minutes, a Diameter Request message has been rejected with an error (Result-Code DIAMETER_UNABLE_TO_DELIVER) in order to protect the system from possible out of memory conditions due to an excess of queued messages. This alarm is cleared automatically if there are no discards due to excessive memory usage for 5 minutes. |
Application | DSC |
Action | See the following Recommended Action |
Excessive memory use for queued requests usually happens because a large number of requests are not answered.
First, check that the DSC MSA objects with Name QUEUE USAGE <slot> have reasonable Max Size attributes. The Maximum Size attribute represents the number of megabytes that may be queued. If this queue size is too low, it could lead to unnecessary request rejection. For more information about DSC MSA objects, refer to Configuring MSAs.
Next, large amounts of queued requests usually mean that requests are timing out. These request timeouts should be generating paced logs in ptidbglog that gives some characteristics of the discarded messages. Investigate the timeouts and see if the root cause can be determined (for example, are all the timeouts relating to the same Destination-Realm and Application-Id). For more information about the ptidbglog, refer to Alarm and Event Logging.
If the alarm persists, contact Customer Support.