feat(failover): add VM 201 manual failover + recovery scripts, watchdog alert

VM 201 (Windows critical) stays out of HA by design. Added:
- failover-vm201.sh: interactive failover pvemini -> pveelite with ZFS replication state
- recover-vm201-to-pvemini.sh: interactive reverse migration with uptime + split-brain checks
- pvemini-down-alert.sh: cron watchdog on pveelite, emails full runbook after 2min DOWN

Replication RPO tightened: CT 108 + VM 201 to 5min, CT 171 to 15min.
CT 171 added to HA (ha-group-main) for continuous Claude Code access.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

proxmox/cluster/failover/failover-vm201.sh

@@ -0,0 +1,79 @@
+#!/bin/bash
+# Manual failover VM 201 (roacentral Windows) către pveelite
+# Deployed la /opt/scripts/failover-vm201.sh pe pveelite
+# Folosește ultima replicare ZFS locală
+set -euo pipefail
+
+VMID=201
+PRIMARY=pvemini
+PRIMARY_IP=10.0.20.201
+SECONDARY=$(hostname)
+LOG=/var/log/failover-vm201.log
+MAIL_TO=mmarius28@gmail.com
+
+log() { echo "[$(date '+%F %T')] $*" | tee -a "$LOG"; }
+die() { log "ABORT: $*"; exit 1; }
+
+[[ "$SECONDARY" == "pveelite" ]] || die "Rulează pe pveelite, nu pe $SECONDARY"
+[[ $EUID -eq 0 ]] || die "Trebuie root"
+
+FORCE=${1:-}
+
+log "=== Failover VM $VMID pornit ==="
+
+# Check 1: pvemini efectiv down
+log "Verific $PRIMARY..."
+if ping -c 2 -W 2 "$PRIMARY_IP" &>/dev/null; then
+    if ssh -o ConnectTimeout=5 -o BatchMode=yes "root@$PRIMARY_IP" 'pvecm status' &>/dev/null; then
+        die "$PRIMARY răspunde la ping + ssh. NU fac failover. Oprește manual VM 201 pe $PRIMARY înainte."
+    fi
+    log "$PRIMARY ping OK dar ssh fail — posibil degradat"
+fi
+log "$PRIMARY confirmat DOWN"
+
+# Check 2: VM nu rulează deja local
+if qm status "$VMID" 2>/dev/null | grep -q running; then
+    die "VM $VMID deja running pe $SECONDARY"
+fi
+
+# Check 3: discul există local (ultima replicare)
+if ! zfs list "rpool/data/vm-$VMID-disk-3" &>/dev/null; then
+    die "Disc vm-$VMID-disk-3 lipsește local. Replicare nefuncțională?"
+fi
+LAST_SNAP=$(zfs list -t snapshot -o name,creation -s creation "rpool/data/vm-$VMID-disk-3" 2>/dev/null | tail -1)
+log "Ultima replicare: $LAST_SNAP"
+
+# Confirmare interactivă
+if [[ "$FORCE" != "--yes" ]]; then
+    echo
+    echo "========================================"
+    echo "ATENȚIE: VM $VMID va porni pe $SECONDARY"
+    echo "folosind ultima replicare ZFS."
+    echo "Date pierdute = ce s-a scris după $LAST_SNAP"
+    echo "========================================"
+    read -p "Continui? (tastează 'DA' pentru confirmare): " CONFIRM
+    [[ "$CONFIRM" == "DA" ]] || die "Anulat de utilizator"
+fi
+
+# Start
+log "Pornesc VM $VMID pe $SECONDARY..."
+qm start "$VMID"
+sleep 5
+STATUS=$(qm status "$VMID" | awk '{print $2}')
+log "Status: $STATUS"
+
+# Mail
+{
+    echo "Failover VM $VMID executat pe $SECONDARY la $(date)"
+    echo
+    echo "Primary $PRIMARY era DOWN."
+    echo "Ultima replicare folosită: $LAST_SNAP"
+    echo "Status VM: $STATUS"
+    echo
+    echo "Pași următori:"
+    echo "1. Verifică aplicația Windows pe 10.0.20.201 (IIS, etc)"
+    echo "2. Când $PRIMARY revine, NU porni VM $VMID pe $PRIMARY"
+    echo "3. După stabilizare, migrare înapoi: qm migrate $VMID $PRIMARY --online 0"
+} | mail -r 'ups@romfast.ro' -s "[CRITIC] Failover VM $VMID pornit pe $SECONDARY" "$MAIL_TO"
+
+log "=== Failover complet ==="