Prometheus metrics reference

Every metric rcfg-sim exposes, with its labels and buckets. Metric names and label keys are public API — see versioning & scope.

Metrics

Metric	Type	Labels	Description
rcfgsim_active_sessions	Gauge	—	SSH sessions currently open across all listeners
rcfgsim_sessions_total	Counter	result	Sessions closed, by terminal result
rcfgsim_session_duration_seconds	Histogram	—	End-to-end session lifetime (accept → close)
rcfgsim_command_duration_seconds	Histogram	command	Per-command dispatch + response latency
rcfgsim_bytes_sent_total	Counter	—	Total bytes written to SSH channels
rcfgsim_auth_attempts_total	Counter	result	SSH password-auth attempts, by result
rcfgsim_handshake_duration_seconds	Histogram	—	TCP accept → completed SSH handshake
rcfgsim_faults_injected_total	Counter	type	Faults injected, by type

Label values

These are the closed sets pre-registered at startup:

• rcfgsim_sessions_total{result} — ok, auth_fail, disconnect, error
• rcfgsim_auth_attempts_total{result} — ok, fail
• rcfgsim_faults_injected_total{type} — auth_fail, disconnect_mid, slow_response, malformed
• rcfgsim_command_duration_seconds{command} — the Cisco command set, pre-registered at zero: CmdUnknown, CmdEmpty, CmdAmbiguous, CmdTerminalLength, CmdTerminalPager, CmdEnable, CmdShowVersion, CmdShowRunningConfig, CmdShowStartupConfig, CmdShowInventory, CmdExit.

TL1 command labels

The Ciena TL1 driver emits additional command values — CmdTL1ActUser, CmdTL1RtrvEqpt, CmdTL1RtrvAlmAll, CmdTL1RtrvCondAll, CmdTL1RtrvActiveUser, CmdTL1RtrvSwVer, CmdTL1RtrvSys, CmdTL1Deny, CmdTL1Unknown. These appear once a Ciena device is served; they are not part of the pre-registered Cisco set.

Histogram buckets

Histogram	Buckets (seconds)
rcfgsim_session_duration_seconds	0.05, 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 30, 60
rcfgsim_command_duration_seconds	0.001, 0.005, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2.5
rcfgsim_handshake_duration_seconds	0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2.5

Runtime & process metrics

Standard Go (go_*) and process (process_*) collectors are also registered — goroutine counts, heap and GC stats, CPU time, and open file descriptors.

See Grafana queries for ready-to-use expressions over these metrics.