From 681d1cebb9a05d3ff0b61cb6110569502aa9076c Mon Sep 17 00:00:00 2001
From: Nicolas Cantu <nicolas.cantu@pm.me>
Date: Tue, 6 Jan 2026 14:40:31 +0100
Subject: [PATCH] Detect and handle ControlSocket errors during SSH execution
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

**Motivations:**
- ControlSocket errors still occurring during command execution
- Socket can become invalid DURING execution of long commands
- Need to detect socket errors in command output and clean up immediately

**Root causes:**
- Socket validation before command execution cannot detect if socket dies during execution
- Long commands (npm install, build) can cause connection to die mid-execution
- Next command finds invalid socket and SSH disables multiplexing but leaves socket
- Previous cleanup only happened before execution, not after detecting errors

**Correctifs:**
- Capture SSH command output to detect 'ControlSocket already exists' errors
- If socket error detected, immediately cleanup and retry once
- This is a specific retry for socket errors only, not a general retry mechanism
- Ensures dead sockets are cleaned up even if they die during command execution

**Evolutions:**
- Better handling of socket invalidation during long-running commands
- Automatic recovery from socket errors detected during execution

**Pages affectées:**
- deploy.sh: Enhanced ssh_exec() to detect and handle socket errors in output
---
 deploy.sh | 29 +++++++++++++++++++++++++----
 1 file changed, 25 insertions(+), 4 deletions(-)

diff --git a/deploy.sh b/deploy.sh
index 944d9a4..ec4a601 100644
--- a/deploy.sh
+++ b/deploy.sh
@@ -45,6 +45,8 @@ check_ssh_connection() {
 ssh_exec() {
     # Toujours vérifier et nettoyer le socket avant chaque commande
     # pour éviter les sockets morts qui causent "ControlSocket already exists, disabling multiplexing"
+    # Le socket peut devenir invalide pendant l'exécution d'une commande précédente,
+    # donc on vérifie systématiquement avant chaque nouvelle commande
     if [ -S "${SSH_CONTROL_PATH}" ]; then
         if ! check_ssh_connection; then
             # Connexion morte, nettoyer avant d'exécuter
@@ -53,15 +55,34 @@ ssh_exec() {
     fi
 
     # Exécuter la commande SSH (une seule tentative, pas de retry)
-    # ControlMaster=auto va créer une nouvelle connexion si le socket n'existe pas,
-    # ou réutiliser s'il est valide
-    ssh -o ControlMaster=auto \
+    # Capture stderr pour détecter les erreurs de socket
+    local ssh_output
+    ssh_output=$(ssh -o ControlMaster=auto \
         -o ControlPath="${SSH_CONTROL_PATH}" \
         -o ControlPersist=300 \
         -o ConnectTimeout=10 \
         -o ServerAliveInterval=60 \
         -o ServerAliveCountMax=3 \
-        ${SERVER} "$@" 2>&1
+        ${SERVER} "$@" 2>&1)
+    local ssh_exit_code=$?
+
+    # Si on détecte une erreur de socket, nettoyer et réessayer une fois
+    if echo "$ssh_output" | grep -q "ControlSocket.*already exists"; then
+        cleanup_dead_ssh
+        # Réessayer une fois après nettoyage
+        ssh -o ControlMaster=auto \
+            -o ControlPath="${SSH_CONTROL_PATH}" \
+            -o ControlPersist=300 \
+            -o ConnectTimeout=10 \
+            -o ServerAliveInterval=60 \
+            -o ServerAliveCountMax=3 \
+            ${SERVER} "$@" 2>&1
+        return $?
+    fi
+
+    # Afficher la sortie et retourner le code de sortie
+    echo "$ssh_output"
+    return $ssh_exit_code
 }
 
 # Nettoyer les connexions SSH persistantes et le répertoire temporaire à la fin