Metrics Reference

Reference for all metrics exposed by Connectware components, including Protocol Mapper, Service Manager, Resource Status Tracking, and Auth Server.

Protocol Mapper, Service Manager, Resource Status Tracking, and Auth Server expose metrics in OpenMetrics format on a dedicated HTTP endpoint at port 9184. This page lists every metric, its type, its labels, and an example scrape output for each component.

For Kubernetes deployments, metrics endpoints and Prometheus Operator ServiceMonitor resources are configured through the connectware and connectware-agent or connectware-agent Helm charts. See Configuring Metrics for Helm-specific configuration.

Common Configuration Parameters

Each component that exposes metrics shares the following environment variables.

Parameter	Default	Description
CYBUS_METRICS_BINDHOST	::	Network interface the metrics HTTP server binds to. The default :: exposes the endpoint on all interfaces. Set to localhost to restrict access to the host only.
CYBUS_METRICS_COLLECTINTERVALMS	5000	How often collected values are stored into Prometheus counters and gauges, in milliseconds. Scrape intervals shorter than this value will return cached data.
CYBUS_METRICS_RATELIMIT_COUNT	1	Maximum number of scrape requests accepted within the rate limit window. When the limit is reached, the cached value for each metric is returned.
CYBUS_METRICS_RATELIMIT_WINDOWMS	1000	Duration of the rate limit window, in milliseconds.

Protocol Mapper

The Protocol Mapper component uses the agent_ metric prefix. Its metrics reflect the current state of connections, endpoints, mappings, nodes, and the overall health of the component.

Metric	Type	Labels	Description
agent_connections_deployed	Gauge	protocol, status	Current number of deployed connections, broken down by protocol and connection status.
agent_endpoints_deployed	Gauge	protocol, operation	Current number of deployed endpoints, broken down by protocol and operation type.
agent_endpoints_messages_total	Counter	protocol, operation	Total number of messages processed by endpoints, broken down by protocol and operation.
agent_mappings_messages_total	Counter	direction	Total number of messages processed by mappings, broken down by direction.
agent_nodes_operations_total	Counter	protocol, operation	Total number of executed node operations, broken down by protocol and operation type.
agent_health	Gauge	component	Health status of the agent components. A value of 1 indicates healthy.

Label values

• agent_endpoints_deployed, agent_endpoints_messages_total — operation: read, write, subscribe

• agent_mappings_messages_total — direction: inbound, outbound

• agent_nodes_operations_total — operation: serverProvides, serverReceives

• agent_health — component: control-plane, data-plane

Example output

agent_connections_deployed{protocol="mqtt",status="connected"} 2
agent_connections_deployed{protocol="opcua",status="connected"} 1

agent_endpoints_deployed{protocol="mqtt",operation="subscribe"} 4
agent_endpoints_deployed{protocol="mqtt",operation="read"} 1
agent_endpoints_deployed{protocol="mqtt",operation="write"} 1
agent_endpoints_deployed{protocol="opcua",operation="subscribe"} 120

agent_endpoints_messages_total{protocol="opcua",operation="subscribe"} 814440
agent_endpoints_messages_total{protocol="mqtt",operation="write"} 1
agent_endpoints_messages_total{protocol="mqtt",operation="read"} 1

agent_mappings_messages_total{direction="inbound"} 4
agent_mappings_messages_total{direction="outbound"} 4

agent_nodes_operations_total{protocol="http",operation="serverProvides"} 1
agent_nodes_operations_total{protocol="http",operation="serverReceives"} 3

agent_health{component="control-plane"} 1
agent_health{component="data-plane"} 1

Service Manager

The Service Manager component uses the sm_ metric prefix. Its metrics track deployed service counts, installation failures, enabling durations, and dependency resolution failures.

Metric	Type	Labels	Description
sm_services_deployed	Gauge	status	Current number of deployed services, broken down by status.
sm_service_installation_failures_total	Counter	—	Total number of service installation failures.
sm_service_enabling_duration_milliseconds	Histogram	—	Duration of service enabling operations, in milliseconds.
sm_unresolved_dependencies_service_enable_failures_total	Counter	—	Total number of service enable failures caused by unresolved dependencies.

Label values

• sm_services_deployed — status: enabled, disabled

Example output

sm_services_deployed{status="enabled"} 0
sm_services_deployed{status="disabled"} 1

sm_service_installation_failures_total 0

sm_service_enabling_duration_milliseconds_bucket{le="25"} 1
sm_service_enabling_duration_milliseconds_bucket{le="50"} 3
sm_service_enabling_duration_milliseconds_bucket{le="100"} 3
sm_service_enabling_duration_milliseconds_bucket{le="250"} 3
sm_service_enabling_duration_milliseconds_bucket{le="1000"} 3
sm_service_enabling_duration_milliseconds_bucket{le="3000"} 3
sm_service_enabling_duration_milliseconds_bucket{le="+Inf"} 3
sm_service_enabling_duration_milliseconds_sum 77
sm_service_enabling_duration_milliseconds_count 3

sm_unresolved_dependencies_service_enable_failures_total 10

Resource Status Tracking

The Resource Status Tracking component uses the rst_ metric prefix. Its metrics report the number of services that have active deviations.

Metric	Type	Labels	Description
rst_services_deviations	Gauge	—	Current number of services with one or more active deviations.

Example output

rst_services_deviations 1

Auth Server

The Auth Server component uses the as_ metric prefix. Its metrics track authentication and authorization failures and request volumes, broken down by request type.

Metric	Type	Labels	Description
as_authentication_failures_total	Counter	type	Total number of authentication failures, broken down by request type.
as_authorization_failures_total	Counter	type	Total number of authorization failures, broken down by request type.
as_authorization_requests_total	Counter	type	Total number of authorization requests processed, broken down by request type.

Label values for type

Value	Applies to	Description
login	as_authentication_failures_total	Failures during the login process.
mfa	as_authentication_failures_total	Failures during multi-factor authentication calls.
http	All three metrics	Any HTTP API call.
mqtt	as_authentication_failures_total	Any MQTT publish or subscribe request.
mqtt_publish	as_authorization_failures_total, as_authorization_requests_total	MQTT publish requests.
mqtt_subscribe	as_authorization_failures_total, as_authorization_requests_total	MQTT subscribe requests.
mqtt5_subscribe	as_authorization_failures_total, as_authorization_requests_total	MQTT 5 subscribe requests.

Example output

as_authentication_failures_total{type="login"} 1
as_authentication_failures_total{type="mqtt"} 2
as_authentication_failures_total{type="http"} 1
as_authentication_failures_total{type="mfa"} 1

as_authorization_failures_total{type="http"} 1
as_authorization_failures_total{type="mqtt_publish"} 1
as_authorization_failures_total{type="mqtt_subscribe"} 1
as_authorization_failures_total{type="mqtt5_subscribe"} 1

as_authorization_requests_total{type="http"} 132
as_authorization_requests_total{type="mqtt_publish"} 13
as_authorization_requests_total{type="mqtt_subscribe"} 12
as_authorization_requests_total{type="mqtt5_subscribe"} 3

Broker and NATS Metrics

Broker and NATS also expose metrics endpoints, but their metrics come from the underlying software rather than from Connectware instrumentation, so this page does not list them.

• Broker metrics are node-level and combine the standard broker metric set with Cybus-specific additions. For the metric list and for scraping a multi-node cluster, see Monitoring & Analytics.

• NATS serves metrics in JSON format. A Prometheus exporter sidecar is deployed by default, which translates NATS metrics into an OpenMetrics compatible format. Setting nats.metrics.enabled to false disables this sidecar. For the available Helm values, see Configuring Prometheus Metrics.

Both components use their own metrics Service and port. For the details, see Metrics Endpoints.