http://jira.hyperic.com/browse/HHQ-4350
RabbitMQ plugin - CPU Usage continues to collect when RabbitMQ Node stopped and unavailable.
After the RabbitMQ server has been added to Hyperic, the CPU Usage shows close to 0% but when the RabbitMQ Node is stopped, CPU Usage jumps to almost 100%. Simply stopping the RabbitMQ Node should not increase the CPU Usage and especialy continue metrics collection.
Trying to reproduce the high CPU Usage, I have seen where the CPU Usage does not show 100% but the problem is still that the metrics continue to collect when the RabbitMQ Node is stopped.
See screenshot 'RabbitMQ CPU Usage.jpg' for visual representation.
Expected Result:
The CPU Usage should remain constant and not collect when RabbitMQ Node stopped
Actual Result:
Stopping the RabbitMQ Node still shows CPU Usage collecting
Steps to Reproduce:
1. Install hq-agent on RabbitMQ platform
2. Log into Hyperic
3. Add RabbitMQ platform to Inventory from Auto-Discovery
4. Select RabbitMQ Node from RabbitMQ platform and configure
5. Allow Availability and Metrics to collect/report
6. Stop the RabbitMQ Node
7. Allow Availability and Metrics to continue collecting/reporting
8. Note the CPU Usage for the RabbitMQ Node continues to collect/report metrics
Additional Information:
Executing top (on the server) when the RabbitMQ Node is stopped shows minimal CPU Usage.
Anonymous