APM Stream Metrics Reference

APM Stream pushes the metrics to a Prometheus Push Gateway instance. You can access the data through the Prometheus platform, as described in Viewing metrics with Prometheus.

All metric messages use the format: <prefix>_<metric_name>{<grouping_keys>} <value> where:

<prefix> is the name of the job representing the source of the metrics. The default value is ta_amp_job.

<metric_name> is the name of the metric being delivered

{<grouping_keys>} is a comma-separated string of key/value pairs used to group metrics based on their values. Each metric has its own set of grouping keys.
<value> is the value of the metric.

In this Queue metric message:

ta_apm_job_queue_running_jobs{instance="",job="ta_apm_job",ta_env_instance="localhost",ta_queue_name="JDE_QUEUE"} 992

ta_apm_job is the prefix.
queue_running_jobs is the name of the metric.
ta_env_instance and ta_queue_name are the two metric-specific grouping keys.
992 is the metric value.

APMJobResource

The APMJobResource message can provide metrics related to virtual resources. For more information about virtual resources, refer to the Resources in TA section in the Tidal Automation User Guide.

Metric Name	Description
<prefix>_virtual_resource_available	Number of remaining virtual resource units a job can consume.
<prefix>_virtual_resource_inuse	Number of virtual resources currently in use for a job execution.
<prefix>_virtual_resource_limit	Maximum virtual resources available for a job.
<prefix>_virtual_resource_locked	Number of virtual resources locked for a job.

APMJobStatus

The APMJobStatus message can provide metrics related to the execution time of a job after its completion.

Metric Name	Description
<prefix>_jobstatusjob_duration_seconds	Actual duration time in seconds.
<prefix>_jobstatus job_failure_count	Number of failed runs for a job instance.
<prefix>_jobstatus job_rerun_count	Number of reruns for a job instance.

APMQueueHealthMessage

The APMQueueHealthMessage message sends the number of messages that have been processed, are currently being processed, and are waiting to be processed in the various TA Master message queues. The TA Master sends this object every five seconds.A set of metrics will report the master queue message thread pool performance.

Metric Name	Description
<prefix>_activity_message_posted	Number of messages posted to the activity queue.
<prefix>_activity_message_processed	Number of messages processed from the activity queue.
<prefix>_activity_message_remain_in_queue	Number of messages remaining in the activity queue.
<prefix>_comm_default_message_posted	Number of messages posted to the default communication queue.
<prefix>_comm_default_message_processed	Number of messages processed from the default communication queue.
<prefix>_comm_default_message_remain_in_queue	Number of messages remaining in the default communication queue.
<prefix>_comm_high_message_posted	Number of messages posted to the high-priority communication queue.
<prefix>_comm_high_message_processed	Number of messages processed from the high-priority communication queue.
<prefix>_comm_high_message_remain_in_queue	Number of messages remaining in the high-priority communication queue.
<prefix>_compiler_message_posted	Number of messages posted to the compiler queue.
<prefix>_compiler_message_processed	Number of messages processed from the compiler queue.
<prefix>_compiler_message_remain_in_queue	Number of messages remaining in the compiler queue.
<prefix>_default_message_posted	Number of messages posted to the default queue.
<prefix>_default_message_processed	Number of messages processed from the default queue.
<prefix>_default_message_remain_in_queue	Number of messages remaining in the default queue.
<prefix>_event_message_posted	Number of messages posted to the event queue.
<prefix>_event_message_processed	Number of messages processed from the event queue.
<prefix>_event_message_remain_in_queue	Number of messages remaining in the event queue.
<prefix>_high_priority_message_posted	Number of posted messages which were used for core thread mission-critical messages were posted.
<prefix>_high_priority_message_processed	Number of messages processed from the high-priority queue.
<prefix>_high_priority_message_remain_in_queue	Number of messages remaining in the high-priority queue.
<prefix>_low_priority_message_posted	Number of messages posted to the low-priority queue.
<prefix>_low_priority_message_processed	Number of messages posted to the low-priority queue.
<prefix>_low_priority_message_remain_in_queue	Number of messages remaining in the low-priority queue.
<prefix>_special_message_posted	Number of messages posted to the special queue.
<prefix>_special_message_processed	Number of messages processed from the special queue.
<prefix>_special_message_remain_in_queue	Number of messages remaining in the special queue.

APMServerResource

The APMServerResource message includes the information related to the performance of the Java Virtual Machine (JVM), SQL database queries, and message queues.

Metric Name	Description
<prefix>_server_resource_average_db_query_time	Reports the average database query time in seconds among 100 calls per polling interval from a TA Master.
<prefix>_server_resource_average_message_queue_time	Reports the average message queuing time in seconds among 100 messages per polling interval from a TA Master.
<prefix>_server_resource_cpu_used	Reports the master JVM process CPU usage of the whole system, in bytes.
<prefix>_server_resource_max_db_query_time	Reports the maximum or longest message queuing time in seconds among 100 messages per polling interval from a TA Master.
<prefix>_server_resource_memory_limit	Reports the assigned maximum heap memory size of a TA Master, in bytes.
<prefix>_server_resource_memory_used	Reports the heap memory usage of a TA Master, in bytes.

Connection

The Connection message includes the information related to the performance of jobs being processed by a TA agent.

Metric Name	Description
<prefix>_connection_job_limit	Maximum number of jobs that can run at the same time on the agent.
<prefix>_connection_load	The current load of the agent.
<prefix>_connection_status	The connection status. 0: Disconnected 1: Active 2: Outage Window (No rebroadcast of metric during the outagewindow) -1: Disabled (No rebroadcast of metric while disabled)

Metric Name

Description

<prefix>_connection_job_limit

Maximum number of jobs that can run at the same time on the agent.

<prefix>_connection_load

The current load of the agent.

<prefix>_connection_status

The connection status.

0: Disconnected
1: Active
2: Outage Window (No rebroadcast of metric during the outagewindow)
-1: Disabled (No rebroadcast of metric while disabled)

Queue

A Queue message can include the information related to the performance of jobs in the TA system.

Metric Name	Description
<prefix>_queue_waiting_jobs	Number of job runs waiting in the queue.
<prefix>_queue_running_jobs	Number of job runs running in the queue.

Statistics

The metrics are provided in the Statistics message pertaining to jobs running in a production day.

Metric Name	Description
<prefix>_stats_jobs_togo	Number of jobs that are waiting and still eligible to run.
<prefix>_stats_jobs_total	Total number of jobs to be run for the current production date.