fixed lock logic. if a continuious app hits a lock it skips to the next instead of error

common.sh

@@ -6,7 +6,7 @@
 # -----------------------------------------------------------------------------------------------
 # Changelog:
 #   v1.0 — Initial stable framework
-#   v1.1 — format_duration moved here from daily_sync.sh for shared use
+#   v1.1 — format_duration moved here from daily_sync_maintenance.sh for shared use
 #           SSH_KEY collision resolved — gitea key renamed GITEA_SSH_KEY in Master.conf
 #           Version and changelog tracking added
 #   v1.2 — Consistent function header comment blocks across all functions
@@ -92,6 +92,7 @@ ICON_STOPPED="🔴"       # container confirmed stopped
 ICON_START="▶️"         # start command being issued
 ICON_STARTED="💚"       # container confirmed started
 ICON_RUNNING="🟢"       # container already running when checked
+ICON_SKIP="⏭️"          # skipping — already running healthy instance
 ICON_NOT_RUNNING="⭕"   # container already stopped when checked
 
 # Transfer
@@ -695,6 +696,8 @@ _release_on_exit() {
 # acquire_lock — acquire exclusive lock for this script
 # Mode: strict (default) — exit immediately if locked
 #       wait             — wait LOCK_WAIT_TIMEOUT seconds then exit
+#       continuous       — for long-running scripts: skip gracefully if healthy,
+#                          clear and restart if dead/stuck
 # Stale lock: if PID in lock file is dead → clear and acquire
 # Age warning: if lock older than LOCK_WARN_AGE → warn
 # -----------------------------------------------------------------------------------------------
@@ -724,7 +727,19 @@ acquire_lock() {
                 warn "$script_name has been running for ${lock_age}s — may be stuck (PID $existing_pid)"
             fi
 
-            if [[ "$mode" == "wait" ]]; then
+            if [[ "$mode" == "continuous" ]]; then
+                # Continuous scripts (watchdogs, failover) — healthy instance = skip gracefully
+                # Only force-restart if PID is stuck/unresponsive beyond LOCK_WARN_AGE
+                if [[ "$lock_age" -gt "$LOCK_WARN_AGE" ]]; then
+                    warn "$script_name appears stuck (running ${lock_age}s) — clearing and restarting"
+                    kill "$existing_pid" 2>/dev/null
+                    sleep 2
+                    rm -f "$lockfile"
+                else
+                    log "$ICON_SKIP $script_name already running healthy (PID $existing_pid) — skipping"
+                    exit 0
+                fi
+            elif [[ "$mode" == "wait" ]]; then
                 info "Another instance of $script_name is running — waiting up to ${LOCK_WAIT_TIMEOUT}s"
                 local waited=0
                 while [[ -f "$lockfile" ]] && [[ "$waited" -lt "$LOCK_WAIT_TIMEOUT" ]]; do
@@ -746,7 +761,7 @@ acquire_lock() {
                 error "Another instance of $script_name is already running (PID $existing_pid) — exiting"
                 # Notify for critical scripts that should rarely overlap
                 case "$script_name" in
-                    failover|transcode_management|media_management|daily_sync|system_watchdog)
+                    failover|transcode_management|media_management|daily_sync_maintenance|system_watchdog)
                         notify "$script_name lock collision on $(hostname) — concurrent instance detected" "$script_name" "warning"
                         ;;
                 esac