Guide
Logs and monitoring
Read the tenraid log, check health over HTTP and scrape Prometheus metrics from the node, the miner, the stratum bridge and the pool.
On this page
Log
| Level | Shows |
|---|---|
| info | Startup summary, listen addresses with the node's --peer value, peers joining and leaving, TenQ dataset builds, new epochs and warnings. Every 10 s the height, new cuts and blocks, waiting transactions and peers. Every 5 s, when there were any, the transactions and blocks submitted to the API |
| debug | Resolved settings, every installed cut with its blocks, transactions and install time, every accepted or rejected block and transaction with the reason, P2P requests |
--log-file <FILE> keeps a copy; RUST_LOG overrides the levels per module.
Health checks
| Request | Answer |
|---|---|
GET /health | ok |
GET /v1/node/sync | synced, finalized_height, best_peer_height, behind_by, peers |
GET /v1/node/status | The node's state at a glance |
Alert on update_required in /v1/node/status: the network follows rules this node's release does not know, and the node stopped until it is updated. update_announced is an earlier hint from a peer that runs a newer release. See Protocol upgrades.
Prometheus
Metrics are off until you give them a listener of their own:
tenraid run --metrics-bind 127.0.0.1The listener answers GET /metrics and nothing else, without authentication. Keep it on loopback or allow only your Prometheus host in the firewall. A scrape never waits for consensus.
| Metric | Type | Meaning |
|---|---|---|
tenrai_node_info{version,network} | gauge | Always 1 |
tenrai_node_uptime_seconds | gauge | Since the start |
tenrai_node_finalized_height | gauge | Newest finalized cut |
tenrai_node_synced, tenrai_node_peers, tenrai_node_best_peer_height | gauge | Sync with the authenticated peers |
tenrai_node_cuts_total, tenrai_node_registered_blocks_total, tenrai_node_executed_blocks_total, tenrai_node_paid_rewards_total | counter | Finalized since the start |
tenrai_node_submissions_total{kind,result} | counter | Transactions and blocks handed to the API |
tenrai_node_mempool_transactions, tenrai_node_work_pool_blocks | gauge | Left out while the engine is busy |
tenrai_node_p2p_received_bytes_total, tenrai_node_p2p_sent_bytes_total | counter | Peer traffic |
tenrai_node_p2p_queued_cuts, tenrai_node_p2p_queued_blocks | gauge | Peer work not applied yet |
tenrai_node_cut_step_seconds{step} | histogram | Time per step of cut installation |
Without a port the listener takes 28790.
Other components
| Component | Option | Metric prefix |
|---|---|---|
tenrai-miner | --metrics-bind 127.0.0.1, port 28793 | tenrai_miner_ |
tenrai-stratum-bridge | --metrics-bind 127.0.0.1, port 28792 | tenrai_bridge_ |
tenrai-pool | /metrics on its web listener, off with --no-metrics | tenrai_pool_ |
The miner and the bridge also serve JSON statistics, /v1/stats, on ports 28770 and 28771. The pool's web listener must stay behind a reverse proxy that keeps /metrics private.