fix(deploy): server-update.sh Health-Check mit Retry (Worker-Startzeit)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

server-update.sh

@@ -58,8 +58,15 @@ state="$(systemctl is-active "$SERVICE" || true)"
 [[ "$state" == "active" ]] || { journalctl -u "$SERVICE" -n 30 --no-pager; die "Service nicht aktiv ($state)."; }
 log "Service aktiv"
 
-# ── 4. Health-Check ───────────────────────────────────────────────────────────
-if curl -sf --max-time 10 "$HEALTH_URL" >/dev/null; then
+# ── 4. Health-Check (mehrere Versuche – Worker brauchen kurz zum Starten) ─────
+health_ok=false
+for attempt in 1 2 3 4 5; do
+    if curl -sf --max-time 10 "$HEALTH_URL" >/dev/null; then
+        health_ok=true; break
+    fi
+    sleep 2
+done
+if $health_ok; then
     log "Health OK ($HEALTH_URL)"
 else
     warn "Health-Check fehlgeschlagen – Logs prüfen: journalctl -u $SERVICE -n 50"