Metrics

How to scrape Prelude TE metrics in Prometheus format, the list of exposed metrics, and additional counters available through the REST API.

Prelude TE exposes a Prometheus-scrapable metrics endpoint with process info, Go runtime stats, BGP peer state, topology counts per domain, output state, and licensing status.

`GET /api/health/metrics`

Returns Prelude TE health and operational metrics in the Prometheus text exposition format (version 0.0.4).

Unlike /api/health, this endpoint is Bearer-protected, like the rest of the API. Create a token under My tokens and pass it in the Authorization header:

curl -fsSL https://te.example.com/api/health/metrics \
  -H "Authorization: Bearer $PRELUDE_TE_TOKEN"

Prometheus scrape config

scrape_configs:
  - job_name: prelude-te
    metrics_path: /api/health/metrics
    scheme: https
    static_configs:
      - targets: ["te.example.com"]
    authorization:
      type: Bearer
      credentials_file: /etc/prometheus/secrets/prelude-te-token

Exposed metrics

All Prelude TE-specific metrics are prefixed with prelude_te_. Standard Go runtime metrics use the conventional go_ prefix.

Process

Metric	Type	Labels	Description
`prelude_te_up`	gauge	—	`1` if the engine is running.
`prelude_te_build_info`	gauge	`version`, `commit`	Build metadata. Value is always `1`; labels carry the info.
`prelude_te_start_time_seconds`	gauge	—	Unix timestamp of process start.
`prelude_te_process_cpu_usage_percent`	gauge	—	Process CPU usage (user+sys) sampled over a 5s window.

Go runtime

Metric	Type	Description
`go_memstats_alloc_bytes`	gauge	Bytes of allocated heap objects currently in use.
`go_memstats_sys_bytes`	gauge	Bytes of memory obtained from the OS.
`go_memstats_heap_inuse_bytes`	gauge	Bytes in in-use heap spans.
`go_goroutines`	gauge	Number of currently running goroutines.

BGP

Metric	Type	Labels	Description
`prelude_te_bgp_peers`	gauge	`state` (`idle`, `connect`, `active`, `opensent`, `openconfirm`, `established`)	Number of configured BGP peers per FSM state.

All six states are emitted on every scrape, including the ones at zero, so your dashboards plot a stable series.

Topology

Metric	Type	Labels	Description
`prelude_te_topology_nodes`	gauge	`domain`	Number of topology nodes per domain.
`prelude_te_topology_links`	gauge	`domain`	Number of topology links per domain.
`prelude_te_topology_prefixes`	gauge	`domain`	Number of topology prefixes per domain.
`prelude_te_topology_updates_total`	counter	`domain`	Total BGP-LS updates processed per domain.
`prelude_te_topology_withdraws_total`	counter	`domain`	Total BGP-LS withdraws processed per domain.

domain is the BGP-LS Instance Identifier — see Domains.

Outputs

Metric	Type	Labels	Description
`prelude_te_outputs`	gauge	`state` (`disconnected`, `connecting`, `connected`, `error`)	Number of outputs per connection state.

All four states are emitted on every scrape.

Licensing

Metric	Type	Description
`prelude_te_license_valid`	gauge	`1` if a license (paid or community) is currently valid, `0` otherwise.
`prelude_te_trial_days_remaining`	gauge	Days remaining on the built-in trial. Negative once expired.

Example response

# HELP prelude_te_up 1 if the topology engine is running.
# TYPE prelude_te_up gauge
prelude_te_up 1
# HELP prelude_te_build_info Build metadata for the running binary.
# TYPE prelude_te_build_info gauge
prelude_te_build_info{version="1.2.3",commit="a1b2c3d"} 1
# HELP prelude_te_bgp_peers Number of configured BGP peers by FSM state.
# TYPE prelude_te_bgp_peers gauge
prelude_te_bgp_peers{state="active"} 0
prelude_te_bgp_peers{state="connect"} 0
prelude_te_bgp_peers{state="established"} 4
prelude_te_bgp_peers{state="idle"} 0
prelude_te_bgp_peers{state="openconfirm"} 0
prelude_te_bgp_peers{state="opensent"} 0
# HELP prelude_te_topology_nodes Number of topology nodes per domain.
# TYPE prelude_te_topology_nodes gauge
prelude_te_topology_nodes{domain="100"} 218
prelude_te_topology_nodes{domain="200"} 100

Useful PromQL

# Engine reachable
up{job="prelude-te"} == 1

# At least one peer established
sum(prelude_te_bgp_peers{state="established"}) > 0

# Any output in error
sum(prelude_te_outputs{state="error"}) > 0

# Topology growth rate (entities per second) over the last 5 minutes
rate(prelude_te_topology_updates_total[5m])

# Trial about to expire
prelude_te_trial_days_remaining < 7 and prelude_te_license_valid == 0

Beyond the metrics endpoint

The endpoint exposes aggregated counters across the whole fleet. For per-entity detail you'd put on an operator dashboard rather than a Prometheus graph, the REST API has it:

Topology stats — GET /api/topology/stats returns per-domain counters with a last-update timestamp. Cheap, and useful as a readiness signal (see Health).
Per-peer counters — each BGP peer carries messages-received, messages-sent, and updates-received counters, reset on every session establishment. Visible on the BGP → Peers list and through /api/bgp/peers/{peerId}.
Per-output state — each output exposes its current state, last-state-change, and last-error through /api/outputs/{outputId}.