Performance events

The following table lists the events that are created for the Performance component.

Table 1. Events for the Performance component
Event	Event Type	Severity	Call Home	Details
perf_api_init_failed	INFO	WARNING	no	Message: Internal API server for performance data could not be started.
				Description: Start of REST API server failed.
				Cause: Start of REST API server failed.
				User Action: Restart monitoring with mmsysmoncontrol restart. Details may be found in /var/adm/ras/mmsysmonitor.log.
pmcollector_down	STATE_CHANGE	ERROR	no	Message: The pmcollector service should be {0}, and is {1}.
				Description: The performance monitor collector is down.
				Cause: This node is in the global configuration of perfmon by using the mmperfmon config show command, but the pmcollector service is not running.
				User Action: Start the service by using the service or systemctl command. Otherwise, remove the node from the global perfmon configuration by using the mmperfmon command.
pmcollector_port_down	STATE_CHANGE	ERROR	no	Message: Performance monitoring collector port {id} ({0}) is not responding.
				Description: One of the performance monitor collector ports is not responding. Check for the affected port in the event message. Queryport and Queryport2 is needed to run queries to get performance data. ipfixport is used to collect metrics and send them to the collector. Federation port is used by collectors to talk to each other.
				Cause: Trying to establish a TCP connection to the local pmcollector port failed. This can be because pmcollector is down or does not respond to network requests.
				User Action: Check the pmcollector process and logs and verify that it runs correctly.
pmcollector_port_up	STATE_CHANGE	INFO	no	Message: Performance monitoring collector port {0} is responding.
				Description: The performance monitor collector port is working.
				Cause: N/A
				User Action: N/A
pmcollector_port_warn	INFO	INFO	no	Message: The pmcollector service port monitor returned an unknown result.
				Description: The monitoring service for performance monitor collector returned an unknown result.
				Cause: Trying to establish a TCP connection to the local pmcollector port failed with an unexpected error.
				User Action: Check the pmcollector process and logs and verify that it runs correctly.
pmcollector_up	STATE_CHANGE	INFO	no	Message: The state of pmcollector service, as expected, is {0}.
				Description: The performance monitor collector is running.
				Cause: N/A
				User Action: N/A
pmcollector_warn	INFO	INFO	no	Message: The pmcollector service has returned an unknown result.
				Description: The monitoring service for performance monitor collector returned an unknown result.
				Cause: The service or systemctl command returned unknown results about the pmcollector service.
				User Action: Check whether the pmcollector is in the expected status by using the service or systemctl command. If there is no pmcollector service, then irrespective of whether the node is in the global mmperfmon config, check the perfmon documentation.
pmsensors_down	STATE_CHANGE	ERROR	no	Message: The pmsensors service should be {0}, but is {1}.
				Description: The performance monitor sensors are down.
				Cause: The node has the perfmon designation in the mmlscluster command, but the pmsensors service is not running.
				User Action: Start the service by using the service or systemctl command. Otherwise, remove the perfmon designation by using the mmchnode command.
pmsensors_up	STATE_CHANGE	INFO	no	Message: The state of pmsensors service, as expected, is {0}.
				Description: The performance monitor sensors are running.
				Cause: N/A
				User Action: N/A
pmsensors_warn	INFO	INFO	no	Message: The pmsensors service returned an unknown result.
				Description: The monitoring service for performance monitor sensors returned an unknown result.
				Cause: The service or systemctl command returned unknown results about the pmsensors service.
				User Action: Check whether the pmsensors is in the expected status by using the service or systemctl command. If there is no pmsensors service, then irrespective of whether the node has the perfmon designation in the mmlscluster command, check the perfmon documentation.