Init with dashboards

Dashboards/README.md

@@ -0,0 +1,23 @@
+# PTP Grafana dashboards
+
+Import each JSON with **Dashboards -> New -> Import** and select your Prometheus datasource.
+
+Included:
+- `ptp-grandmaster.json` — MOMI / GNSS-PPS / Chrony / PTP
+- `ptp-linux-boundary.json` — WOLF-HUNTER-style Linux boundary clock
+- `ptp-linux-client.json` — selectable downstream Linux client
+- `ptp-nyater.json` — NYATER Nexus boundary clock
+- `ptp-homelab-overview.json` — whole timing-chain overview
+
+The dashboards discover instances from:
+- `ptp_linux_node_info{role="grandmaster"}`
+- `ptp_linux_node_info{role="boundary"}`
+- `ptp_linux_node_info{role="client"}`
+- `ptp_nexus_target_info`
+
+Notes:
+- The GM dashboard does **not** treat `ptp_gm_present=0` as a fault, because the local clock itself is the GM.
+- The WOLF dashboard gives special attention to `ptp_phc2sys_*`, useful for your JBOD boundary-clock PHCs.
+- The NYATER dashboard shows slave/passive port state and Delay_Req migration, which makes redundant-link failover visible.
+- Packet-rate panels use Grafana's `$__rate_interval`.
+- If `PORT_PROPERTIES_NP` is unavailable, Linux port counters may show PTP port identity instead of the friendly interface name.