Skip to content

USER GUIDE

CLI: events

Purpose

Tail normalized carrier, routable, routes, and manager events from the daemon API — the same stream hooks react to, but for operators debugging link flaps or missing scripts.

When to use it

  • A hook did not fire and you need to see whether netevd saw the event
  • You are validating filters in /etc/netevd/netevd.yaml while bouncing an interface
  • You want a live feed during a change window instead of parsing journalctl

How to get there

  • Command: netevd events
  • Default API: http://127.0.0.1:9090 (override with --endpoint http://<host>:9090)
  • Requires api.enabled: true in /etc/netevd/netevd.yaml and a running netevd.service

Operate from CLI

  1. Confirm the daemon and API are up:
systemctl is-active netevd
curl -sf http://127.0.0.1:9090/health | jq .
  1. Show the last 20 events (text table):
netevd events --tail 20
  1. Follow live events (like tail -f):
netevd events -f
  1. Filter by interface and event type while testing a link bounce:
netevd events -f -i eth0 -t routable
  1. Machine-readable output for automation:
netevd events --tail 50 -f json --endpoint http://<host>:9090
  1. Cross-check hook side effects in the journal:
journalctl -u netevd -f

  1. Empty / fail: API unreachable → check api.enabled, bind address, and firewall; events list empty → confirm system.backend matches the host (systemd-networkd, NetworkManager, or dhclient) and the interface is not ignored by a YAML filter.

  2. Success: Events appear within ~100 ms of a link or address change; filtered -i/-t flags narrow the stream to the interface under test.