Metrics Reference

A Demos node ships a built-in Prometheus exporter. When enabled, a dedicated HTTP server serves the standard GET /metrics endpoint (and a simple /health / /healthz returning {"status":"ok"}) on its own port, separate from the RPC server. Every metric name is prefixed with demos_. In addition to the metrics below, the exporter also emits the default prom-client process/Node.js metrics (memory, CPU, event loop, GC), which are also demos_-prefixed.

Enabling

Metrics are controlled by the metrics config block:

Config key	Default	Description
`metrics.enabled`	`true`	Master switch for the exporter.
`metrics.port`	`9090`	Port the metrics HTTP server binds.
`metrics.host`	`0.0.0.0`	Bind address.

See Node Configuration for how these map to environment variables and the Docker Compose / Grafana stack that scrapes this endpoint. Collected gauges are refreshed on a 2.5s interval; counters are incremented at their call sites.

Core metrics

System

Metric	Type	Labels
`demos_node_uptime_seconds`	gauge	—
`demos_node_info`	gauge	`version`, `node_id`

Consensus and chain

Metric	Type	Labels / Buckets
`demos_consensus_rounds_total`	counter	—
`demos_consensus_round_duration_seconds`	histogram	buckets: `0.1, 0.5, 1, 2, 5, 10, 30`
`demos_block_height`	gauge	—
`demos_mempool_size`	gauge	—

Network and peers

Metric	Type	Labels / Buckets
`demos_peers_connected`	gauge	—
`demos_peers_total`	gauge	—
`demos_messages_sent_total`	counter	`type`
`demos_messages_received_total`	counter	`type`
`demos_peer_latency_seconds`	histogram	buckets: `0.01, 0.05, 0.1, 0.25, 0.5, 1, 2.5, 5` (aggregated, no per-peer label)

Transactions

Metric	Type	Labels / Buckets
`demos_transactions_total`	counter	`type`, `status`
`demos_transactions_failed_total`	counter	`type`, `reason`
`demos_tps`	gauge	—
`demos_transaction_processing_seconds`	histogram	`type`; buckets: `0.001, 0.005, 0.01, 0.05, 0.1, 0.5, 1, 5`

API

Metric	Type	Labels / Buckets
`demos_api_requests_total`	counter	`method`, `endpoint`, `status_code`
`demos_api_request_duration_seconds`	histogram	`method`, `endpoint`; buckets: `0.001, 0.005, 0.01, 0.05, 0.1, 0.25, 0.5, 1, 2.5, 5, 10`
`demos_api_errors_total`	counter	`method`, `endpoint`, `error_code`

IPFS and GCR

Metric	Type	Labels
`demos_ipfs_pins_total`	gauge	—
`demos_ipfs_storage_bytes`	gauge	—
`demos_ipfs_peers`	gauge	—
`demos_ipfs_operations_total`	counter	`operation`
`demos_gcr_accounts_total`	gauge	—
`demos_gcr_total_supply`	gauge	—

Observability metrics

These mirror the subsystem registry and boot tracker that also back the /health endpoint.

Metric	Type	Labels	Notes
`demos_subsystem_up`	gauge	`subsystem`	`1` when status is `ready`/`running`, else `0`.
`demos_subsystem_skipped`	gauge	`subsystem`	`1` when intentionally skipped.
`demos_subsystem_uptime_seconds`	gauge	`subsystem`	Seconds since the last state transition.
`demos_subsystem_restart_total`	counter	`subsystem`	Transitions back to `running` after `failed`/`ready`.
`demos_subsystem_error_total`	counter	`subsystem`	Errors recorded via `subsystemError()`.
`demos_boot_step_status`	gauge	`step`	Enum: `0`=pending, `1`=running, `2`=ready, `3`=failed, `4`=skipped.
`demos_boot_complete`	gauge	—	`1` when all boot steps are ready/skipped.
`demos_dormant_mode`	gauge	—	`1` when the node booted with no peers.
`demos_main_loop_heartbeat_seconds`	gauge	—	Seconds since the last main-loop tick.
`demos_main_loop_iterations_total`	counter	—	Main-loop iterations since boot.
`demos_port_drift`	gauge	`subsystem`	Actual minus requested port (`0` normally).
`demos_uncaught_exception_total`	counter	`source`	`uncaughtException` handler invocations.
`demos_unhandled_rejection_total`	counter	`source`	`unhandledRejection` handler invocations.

Collector metrics

Sampled live from node state and the host on each collection tick.

Block detail

Metric	Type	Labels
`demos_last_block_tx_count`	gauge	—
`demos_seconds_since_last_block`	gauge	—
`demos_last_block_timestamp`	gauge	—

Host system

Metric	Type	Labels
`demos_system_cpu_usage_percent`	gauge	—
`demos_system_memory_used_bytes`	gauge	—
`demos_system_memory_total_bytes`	gauge	—
`demos_system_memory_usage_percent`	gauge	—
`demos_system_load_average_1m`	gauge	—
`demos_system_load_average_5m`	gauge	—
`demos_system_load_average_15m`	gauge	—

Network I/O (per interface)

Metric	Type	Labels
`demos_system_network_rx_bytes_total`	gauge	`interface`
`demos_system_network_tx_bytes_total`	gauge	`interface`
`demos_system_network_rx_rate_bytes`	gauge	`interface`
`demos_system_network_tx_rate_bytes`	gauge	`interface`

Service and peer health

Metric	Type	Labels
`demos_service_docker_container_up`	gauge	`container`
`demos_service_port_open`	gauge	`port`, `service`
`demos_peer_online_count`	gauge	—
`demos_peer_offline_count`	gauge	—
`demos_peer_info`	gauge	`peer_id`, `url`, `status` (capped at 20 peers)
`demos_node_http_health`	gauge	`endpoint` (`1`=responding)
`demos_node_http_response_time_ms`	gauge	`endpoint`
`demos_node_metadata`	gauge	`version`, `version_name`, `identity` (always `1`)

Sample scrape

# HELP demos_block_height Current block height
# TYPE demos_block_height gauge
demos_block_height 184213

# HELP demos_subsystem_up Whether the named subsystem is in 'ready' or 'running' state (1) or not (0)
# TYPE demos_subsystem_up gauge
demos_subsystem_up{subsystem="chain"} 1
demos_subsystem_up{subsystem="rpc"} 1
demos_subsystem_up{subsystem="mcp"} 0

# HELP demos_boot_complete 1 when every registered boot step is in ready/skipped, 0 otherwise
# TYPE demos_boot_complete gauge
demos_boot_complete 1

​Metrics Reference

​Enabling

​Core metrics

​System

​Consensus and chain

​Network and peers

​Transactions

​API

​IPFS and GCR

​Observability metrics

​Collector metrics

​Block detail

​Host system

​Network I/O (per interface)

​Service and peer health

​Sample scrape