BCP POWER FAILURE BONIXER 88/100 SOVEREIGN 7 DIAMONDS T+15m RECOVERY 6 SILOS + AKS Day 112 γ₁ = 14.134725141734693  ·  LABR-BCP-POWER-FAILURE-V14-001
SECTION 1 — FLEET BCP SCORE DAY 112 STATE UP FROM 78 → 88
BCP progression: Day 104 92/100 → Day 109 78/100 (degraded) → Day 112 88/100 (recovering)
Service Layer 36 / 40
hermes/portal/mefine/mal-router live · -4: kali restart policy, RSV, mal-router tier, eose-dev bare
Data Layer 25 / 25
NAS + git golden tags + yone qdrant — all intact
Config Layer 18 / 20
rc.local + systemd + portproxy — -2 iphlpsvc risk (D-BCP-PWR-003)
Change Control 9 / 15
ARB1 + sorry-flow active — -6 DR untested (Azure RSV empty)
SCORE HISTORY
Day 104
92/100
SOVEREIGN — pre-Day 109 disruption
Day 109
78/100
DEGRADED — iphlpsvc failure + RSV gap
Day 112
88/100
RECOVERING — rc.local fix deployed
SECTION 2 — SILO BCP GRID 6 SILOS + AKS CLOUD
msi01 ⚓ Admiral Builder 192.168.2.18
RTX 5090 24GB  ·  WSL2 utpemos  ·  Windows 11
✅ AUTOMATIC rc.local + 14 systemd units
Auto-restart: openclaw-gw · hermes-gw · pemos-portal · mefine · laam-pip · psgraph · context-triad · pemclau · euler · lucien-mesh · joffe-v1 · joffe-v2 · now-playing
Manual T+12m: portproxy IP check · Tailscale serve · iphlpsvc verify
⚠️ iphlpsvc must be Running (kills portproxy if stopped)
87/100
T+2m auto · T+12m verified
msclo ⚖️ Admiral Law 192.168.2.19
RTX 5090 24GB  ·  WSL2 + Docker  ·  Windows 11
✅ AUTOMATIC 14 systemd + Docker restart:unless-stopped
Auto-restart: openclaw-gw · laam-pip · hermes-gw · mal-router · context-triad · loom-dual-rail · k3d-lhvcp · cancan-moss · cancan-marb
Manual: mal-router tier IPs if stale after restart
85/100
T+5m auto · T+7m verified
yone γ₁ Validator 192.168.2.23
RTX 5080 16GB  ·  Docker restart:unless-stopped  ·  Windows 11
✅ AUTOMATIC Docker restart:unless-stopped
Auto-restart: qdrant · redis · ollama · hermes · portal · pemclau-sessions · lhvcp
Manual: verify Tailscale 100.78.58.67 reachable post-boot
90/100
T+7m
forge 🔥 Engines 192.168.2.12
RTX 4090 24GB  ·  Docker restart:always (full stack)  ·  Windows 11
✅ AUTOMATIC Docker restart:always — full 24+ container stack
Auto-restart: 24+ containers — pemos stack · laam · redis · qdrant · neo4j · ollama · psgraph-bridge · mevault-local · joffe-math
Manual: none — fully automatic
✅ Most reliable restart in the fleet
95/100
T+8m
lounge 💡 Compute 192.168.2.29
RTX 5090 32GB  ·  WSL utpemos + Docker  ·  Windows 11
⚠️ PARTIAL WSL utpemos + k3d auto · kali containers ❌
Auto-restart: k3d-lhvcp · joffe-math services :9383/:9384/:9385 · utfoundry
Manual T+10m: docker start kali-anubis kali-horus
❌ kali containers missing restart flag — D-BCP-PWR-007
75/100
T+10m partial
eose-dev 📐 QE Backstop 192.168.2.21
RTX A2000  ·  BARE METAL LINUX (not WSL) · Full systemd
✅ AUTOMATIC bare metal Linux — full systemd — fastest restart
Auto-restart: k3s · ollama · hermes-gw · qdrant · redis · neo4j · sorry-tracker
Manual: none — Linux auto-starts all services
✅ Fastest restart in fleet — no Windows boot delay
92/100
T+3m fastest
AKS ☁️ ALWAYS LIVE
pemos.ca  ·  master-system + master1-system
Azure cloud — NOT affected by local power failure. pemos.ca serves traffic while local fleet recovers. This IS the backstop.
Services live: portal · hermes · mefine API · mal-router cloud tier · PEMCLAU GraphRAG endpoint
99/100
-1pt: RSV empty
T+0m ALWAYS UP
SECTION 3 — POWER RECOVERY TIMELINE T+15m = FLEET SOVEREIGN
OPEN DIAMONDS 7 ITEMS 3 RED · 4 AMBER
◆ D-BCP-PWR-001
WSL IP changes portproxy on restart
⚠️ AUTO-MITIGATED — rc.local refresh script deployed msi01
◆ D-BCP-PWR-002
Tailscale route override (kills LAN 192.168.2.0/24)
⚠️ AUTO-MITIGATED — rc.local route fix + serve restart
◆ D-BCP-PWR-003
iphlpsvc Windows service stops after update/reboot
❌ MANUAL CHECK — kills all portproxy (80, 18799, 2225, etc.)
◆ D-BCP-PWR-004
Azure RSV (Recovery Services Vault) empty — no cloud backup
❌ NO DR — AKS deployment covers but data loss possible if NAS fails simultaneously
◆ D-BCP-PWR-005
yone LAN IP drift (DHCP renewal)
⚠️ MITIGATED — use Tailscale 100.78.58.67 path as primary
◆ D-BCP-PWR-006
eose-dev UPS status unknown
⚠️ COVERED — AKS backstop absorbs if eose-dev has dirty shutdown
◆ D-BCP-PWR-007
kali-anubis + kali-horus missing Docker restart flag
❌ MANUAL START — docker start kali-anubis kali-horus needed at T+10m
MANUAL ACTION SCRIPT T+12m msi01 STEPS
# ─── T+12m: msi01 Manual BCP Steps (run as Admin in PowerShell) ──────────────
# STEP 1: Verify iphlpsvc (kills portproxy if stopped)
Get-Service iphlpsvc
# If stopped:
Restart-Service iphlpsvc -Force
Set-Service iphlpsvc -StartupType Automatic

# STEP 2: Verify/fix portproxy (WSL IP changes on restart)
netsh interface portproxy show all
# If stale WSL IP (wrong 172.x.x.x):
$wslIP = (wsl -d utpemos hostname -I).Trim().Split(' ')[0]
netsh interface portproxy delete v4tov4 listenport=80 listenaddress=0.0.0.0
netsh interface portproxy add v4tov4 listenport=80 listenaddress=0.0.0.0 connectaddress=$wslIP connectport=80
netsh interface portproxy add v4tov4 listenport=18799 listenaddress=0.0.0.0 connectaddress=$wslIP connectport=18799
netsh interface portproxy add v4tov4 listenport=2225 listenaddress=0.0.0.0 connectaddress=$wslIP connectport=2225

# STEP 3: Tailscale Serve restore
tailscale serve --bg http://127.0.0.1:80

# ─── WSL: verify Tailscale route ────────────────────────────────────────────
ip route show | grep 192.168.2
# Should NOT have: 192.168.2.0/24 via 100.100.100.100
# If broken: sudo ip route del 192.168.2.0/24 via 100.100.100.100

# ─── lounge: kali containers ────────────────────────────────────────────────
docker start kali-anubis kali-horus

# ─── T+15m: Run BCP bonixer to confirm ≥90/100 ──────────────────────────────
curl -s http://pemos.ca/bcp-power-bonixer | head -5