diff --git a/scripts/clear-all-transaction-pools.sh b/scripts/clear-all-transaction-pools.sh
index ee21175..ca9cf50 100755
--- a/scripts/clear-all-transaction-pools.sh
+++ b/scripts/clear-all-transaction-pools.sh
@@ -1,6 +1,10 @@
 #!/usr/bin/env bash
 # Clear transaction pools on all Besu nodes (RPC and Validators)
 # This script clears transaction pool databases to remove stuck transactions
+#
+# SSH: uses PROXMOX_SSH_USER from config/ip-addresses.conf (root). If .env sets PROXMOX_USER=root@pam for the API,
+# that value is NOT used for SSH (see PROXMOX_USER= assignment below).
+# Afterward: Core RPC (2101) can take several minutes to bind JSON-RPC while RocksDB opens/compacts; public RPC may recover first.
 
 set -euo pipefail
 
@@ -9,10 +13,12 @@ SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 PROJECT_ROOT="$(cd "$SCRIPT_DIR/.." && pwd)"
 source "${PROJECT_ROOT}/config/ip-addresses.conf" 2>/dev/null || true
 
-
-PROXMOX_USER="${PROXMOX_USER:-root}"
-PROXMOX_ML110="${PROXMOX_ML110:-192.168.11.10}"
-PROXMOX_R630="${PROXMOX_R630:-192.168.11.11}"
+# Shell SSH must be root@host — not root@pam@host (.env often sets PROXMOX_USER=root@pam for API).
+PROXMOX_SSH_USER="${PROXMOX_SSH_USER:-root}"
+[[ "$PROXMOX_SSH_USER" == *"@"* ]] && PROXMOX_SSH_USER="root"
+PROXMOX_USER="$PROXMOX_SSH_USER"
+PROXMOX_ML110="${PROXMOX_ML110:-${PROXMOX_HOST_ML110:-192.168.11.10}}"
+PROXMOX_R630="${PROXMOX_R630:-${PROXMOX_R630_01:-${PROXMOX_HOST_R630_01:-192.168.11.11}}}"
 
 # Colors
 RED='\033[0;31m'
diff --git a/scripts/deployment/preflight-chain138-deploy.sh b/scripts/deployment/preflight-chain138-deploy.sh
index e85ce82..fa3159b 100755
--- a/scripts/deployment/preflight-chain138-deploy.sh
+++ b/scripts/deployment/preflight-chain138-deploy.sh
@@ -4,6 +4,8 @@
 #
 # Usage: ./scripts/deployment/preflight-chain138-deploy.sh [--cost]
 #   --cost  Also run cost estimate (smom-dbis-138/scripts/deployment/calculate-costs-consolidated.sh).
+# Optional env: CHAIN138_PREFLIGHT_RPC_URL=https://...  Same chainId 138; use when Core RPC (2101) is down briefly
+#   (e.g. after tx-pool clear / RocksDB WAL) and public RPC (e.g. 192.168.11.221:8545) is up. Deploy still use Core when healthy.
 #
 # Requires: run from repo root. Uses smom-dbis-138/.env only (PRIVATE_KEY, RPC_URL_138).
 # See: docs/03-deployment/DEPLOYMENT_ORDER_OF_OPERATIONS.md § Deployment safety
@@ -54,6 +56,11 @@ else
 fi
 
 RPC="${RPC_URL_138:-http://192.168.11.211:8545}"
+# Optional: same chain, different endpoint (e.g. public RPC 2201) when Core 2101 is restarting after pool maintenance.
+if [[ -n "${CHAIN138_PREFLIGHT_RPC_URL:-}" ]]; then
+  RPC="$CHAIN138_PREFLIGHT_RPC_URL"
+  echo "NOTE: Using CHAIN138_PREFLIGHT_RPC_URL for this preflight only: $RPC" >&2
+fi
 if [[ -z "${PRIVATE_KEY:-}" ]]; then
   echo "FAIL: PRIVATE_KEY not set in $SMOM/.env." >&2
   exit 1