docs: HA VM lost IPv4 — root cause of months-long TV outage; pinned .146

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

.claude/handoffs/2026-10-22-ha-lost-ipv4.md

@@ -0,0 +1,31 @@
+# Handoff: "hasn't worked in months" = HA VM had no IPv4
+
+## Root cause
+HA VM 706 (pve173) was running but had **no IPv4 on enp0s18** — only IPv6 (ULA
+fd70:582a:b527:d4c8:fdaf:a8f9:f05a:40b6). DHCP method `auto`, empty address list.
+Uptime 57d at discovery → lease lost around late Aug 2026; why it never renewed is
+**unknown**. Effect: HA couldn't reach Mosquitto (.154), Caddy couldn't reach .146,
+Google Home → HA dead. The lgtv bridge on steel141 was healthy the whole time
+(polling correctly, connected to broker) — it just never received a command.
+
+## Fix (2026-10-22)
+1. Omada reservation: `network-tools/omada_reserve_ip.py reserve BC:24:11:3F:DC:7B 192.168.0.146`
+2. Re-ran DHCP from HA (`ha network update enp0s18 --ipv4-method auto`) → got .43
+   (router kept the existing lease; one disable/auto cycle didn't change it).
+3. Set HAOS static: `ha network update enp0s18 --ipv4-method static --ipv4-address
+   192.168.0.146/24 --ipv4-gateway 192.168.0.1 --ipv4-nameserver 192.168.0.1`.
+4. `ha core restart`.
+
+## Verified
+- .146 pings; hass.sethpc.xyz → 200.
+- `switch.big_tv_lg_tv` = off (matches TV) → MQTT path is live.
+
+## Not verified / open
+- **Seth to test**: "Hey Google, turn on the TV" (not done by me at 23:00).
+- Projector `media_player.mitsubishi_projector` unavailable — projector itself is
+  off the network (PJLink :4352 closed, inactive in Omada). Probably powered off at
+  the wall; separate from this outage.
+- `media_player.lg_webos_tv_cd04` unavailable is expected while the TV is off.
+- Stale .43 lease will just expire.
+- Recovery path if HA goes dark again: reach it over IPv6 ULA above
+  (`ssh seth@<ULA>`, `bash -lc "ha network info"`).

DECISIONS.md

@@ -47,6 +47,8 @@ Format: `YYYY-MM-DD: <decision> — <why>`
   *not* part of smart-home linking — don't touch it (shared project runs WP Mail SMTP).
   Full detail + security-cleanup list in `.claude/handoffs/2026-06-06-*.md`.
 
+- 2026-10-22: HA VM 706 pinned to 192.168.0.146 twice over — Omada DHCP reservation (MAC BC:24:11:3F:DC:7B) AND HAOS static IPv4 (`ha network update enp0s18 --ipv4-method static`) — HA had silently lost its IPv4 lease (~Aug 2026; IPv6-only, uptime 57d), cutting it off from MQTT/Caddy/Google: the "TV hasn't worked in months" outage. Static removes the DHCP dependency; reservation keeps the router from handing .146 to anyone else.
+
 ## Deferred / Rejected
 - **DHCP reservation for the TV (static IP):** not done — Pi-hole doesn't run
   DHCP (router does), so it's Seth's action on the router. The cached-IP probe