Interpreting object queue statistics from the Statistics page
You can use object queue statistics on the Statistics page in NexJ System Admin Console to view aggregate object queue metrics by node or by queue.
You can view aggregate values for all queues by navigating to and for individual queues stats by navigating to and then selecting a queue node.
Aggregate object queue metrics by node
You can access object queue statistics for your NexJ application for individual nodes or the cluster. You can access these statistics on the Object Queue Engine node.
- Dispatcher
- Displays statistics for the active object queue dispatcher node and measures the performance of the object queue engine
- Persistent
- Displays statistics for messages that are saved to OQMessage table in the process of delivery. Persistent messages typically contain business logic.
- Transient
- Displays statistics for messages that are not saved to a data source. Transient message are typically internal system notifications.
- Sender
- Displays statistics for messages that are on a queue.
- Receiver
- Displays statistics for messages that have been processed and taken off a queue.
Object queue metrics by queue
The following table identifies statistics that you can use to monitor object queue statistics and to review error and runtime queues. You can access these statistics by navigating to the node and then selecting a channel.
| Statistic | Description | How to interpret |
|---|---|---|
| Total message count | The total number of messages that were sent to and received from the queue | Use the difference between Receiver and Sender message counts to determine the number of messages not yet processed. When a message fails once and then succeeds, the send count increments once and the receive count increments twice. |
| Processing time (ms): mean | Receiver node: The average processing time of a successful message Sender node: The average time to add a message to a queue |
Track as a trend. Compare the mean processing time to the universal processing time and look for large changes in processing time from the norm. Also look for large mean processing times, especially where they deviate from the universal processing time. |
| Message rate (Hz) | Receiver node: The rate at which messages were received by an object queue and processed Sender node: The rate at which messages were sent to be received by an object queue |
Track as a trend. |