Skip to main content

First look

sudo nomctl status
Service   go-zenon active (running), pid 1234, 0 restarts, up 3d 4h
Node      znnd v0.0.7 (a1b2c3d), syncing 1,234,567 / 2,000,000 (61.7%), 5.2 mom/s, ETA 1d 16h
Peers     14 connected
Frontier  height 1,234,567, 3s ago
Pillar    MyPillar rank 12, produced 118 / 121 expected this epoch, 3 missed
Process   cpu 42.0%, rss 1.9 GiB, threads 38, open files 412 / 32768
Host      load 1.20 0.90 0.80, mem 3.1 GiB / 7.8 GiB available, /root/.znn 210.0 GiB free
Pressure  cpu 2.1%, io 15.4%, mem 0.0%

CPU % and the sync rate are measured over --wait (default 2 seconds). --json prints the raw sample for scripts.

  • Service: active (running) with a restart count that is not climbing is healthy. A growing count is a crash loop; collect a support bundle with --watch.
  • Node: syncing with a rate above zero means progress. synced with a frontier older than two minutes is shown as [STALLED]; restart the service. not enough peers usually means port 35995/TCP is blocked inbound or the host has no outbound connectivity.
  • Frontier: the newest momentum the node has. Momentums arrive roughly every 10 seconds, so on a healthy synced node this is always a few seconds old.
  • Pillar (only when a pillar name is configured): produced should track expected through the epoch; a growing gap means missed production slots.
  • node rpc unreachable: the process is not up, or its HTTP RPC on port 35997 was disabled in config.json. The rest of the output is still valid.
  • Process: open files near the 32768 limit, or memory close to the host total, predict the two most common crashes.
  • Host: less than 15 GB free on the data directory stops backups and will eventually stop the node. IO pressure above about 50% on a syncing node means the disk is the bottleneck.

Live​

sudo nomctl top

The same values refreshed every two seconds (--interval), with a sync progress bar and a sparkline of momentums per second. q quits. Both commands are read-only and need no configuration.