Skip to content

Metrics

The argocd-agent exposes different sets of Prometheus metrics for agent and principal components.

Metrics are by default enabled in both principal and agent but they will be disabled if metrics port is set to 0.

Metrics for principal are exposed at http://0.0.0.0:8000/metrics endpoint. Users can overwrite metrics port by setting --metrics-port flag in CLI or ARGOCD_PRINCIPAL_METRICS_PORT environment variable.

Similarly for agent metrics are exposed at http://0.0.0.0:8181/metrics endpoint and port can be overwritten by the setting --metrics-port flag in CLI or ARGOCD_AGENT_METRICS_PORT environment variable.

Here is the list of available metrics:

Principal Metrics

Metric Type Description
agent_connected_with_principal gauge The total number of agents connected with principal.
principal_agent_avg_connection_time gauge The average time all agents are connected for (in minutes).
principal_applications_created counter The total number of applications created by agents.
principal_applications_updated counter The total number of applications updated by agents.
principal_applications_deleted counter The total number of applications deleted by agents.
principal_app_projects_created counter The total number of app project created by agents.
principal_app_projects_updated counter The total number of app project updated by agents.
principal_app_projects_deleted counter The total number of app project deleted by agents.
principal_events_received counter The total number of events sent by principal.
principal_events_sent counter The total number of events sent by principal.
principal_event_processing_time histogramVec Histogram of time taken to process events (in seconds).
principal_errors counterVec The total number of errors occurred in principal.

Agent Metrics

Metric Type Description
agent_events_received counter The total number of events received by agent.
agent_events_sent counter The total number of events sent by agent.
agent_event_processing_time histogramVec Histogram of time taken to process events (in seconds).
agent_errors counterVec The total number of errors occurred in agent.

Here is the list of available labels:

Labels

Label Name Example Value Description
call_status success Status of event processing. Possible values are: success, failure, discarded, not-allowed.
agent_name agent-managed Name of Agent. Possible values are: agent-managed, agent-autonomous.
resource_type application Type of resource. Possible values are: application, app project, resource, resourceResync.