ProxmoxVE/misc/error_handler.func

#!/usr/bin/env bash
# ------------------------------------------------------------------------------
# ERROR HANDLER - ERROR & SIGNAL MANAGEMENT
# ------------------------------------------------------------------------------
# Copyright (c) 2021-2026 community-scripts ORG
# Author: MickLesk (CanbiZ)
# License: MIT | https://github.com/community-scripts/ProxmoxVE/raw/main/LICENSE
# ------------------------------------------------------------------------------
#
# Provides error handling and signal management for all scripts.
#
# TELEMETRY CONTRACT:
#   - HOST context: this file reports terminal statuses via post_update_to_api
#     (full metadata + focused error trace).
#   - CONTAINER context: this file NEVER talks to the telemetry API. It writes
#     two local artifacts that the host picks up after lxc-attach returns:
#       /root/.install-<SESSION_ID>.failed        → exit code (flag file)
#       /root/.install-<SESSION_ID>.log.errinfo   → structured error capture
#     This guarantees the server only ever sees ONE terminal event per
#     execution - the host's complete one.
#
# Usage:
#   source <(curl -fsSL .../error_handler.func)
#   catch_errors
#
# ------------------------------------------------------------------------------

# ==============================================================================
# SECTION 1: EXIT CODE EXPLANATIONS (fallback)
# ==============================================================================

# ------------------------------------------------------------------------------
# explain_exit_code()
#
# - Canonical version lives in api.func (sourced before this file on the host)
# - This fallback covers the container context where api.func is not sourced
# ------------------------------------------------------------------------------
if ! declare -f explain_exit_code &>/dev/null; then
  explain_exit_code() {
    local code="$1"
    case "$code" in
    1) echo "General error / Operation not permitted" ;;
    2) echo "Misuse of shell builtins (e.g. syntax error)" ;;
    3) echo "General syntax or argument error" ;;
    10) echo "Docker / privileged mode required (unsupported environment)" ;;
    4) echo "curl: Feature not supported or protocol error" ;;
    5) echo "curl: Could not resolve proxy" ;;
    6) echo "curl: DNS resolution failed (could not resolve host)" ;;
    7) echo "curl: Failed to connect (network unreachable / host down)" ;;
    8) echo "curl: Server reply error (FTP/SFTP or apk untrusted key)" ;;
    16) echo "curl: HTTP/2 framing layer error" ;;
    18) echo "curl: Partial file (transfer not completed)" ;;
    22) echo "curl: HTTP error returned (404, 429, 500+)" ;;
    23) echo "curl: Write error (disk full or permissions)" ;;
    24) echo "curl: Write to local file failed" ;;
    25) echo "curl: Upload failed" ;;
    26) echo "curl: Read error on local file (I/O)" ;;
    27) echo "curl: Out of memory (memory allocation failed)" ;;
    28) echo "curl: Operation timeout (network slow or server not responding)" ;;
    30) echo "curl: FTP port command failed" ;;
    32) echo "curl: FTP SIZE command failed" ;;
    33) echo "curl: HTTP range error" ;;
    34) echo "curl: HTTP post error" ;;
    35) echo "curl: SSL/TLS handshake failed (certificate error)" ;;
    36) echo "curl: FTP bad download resume" ;;
    39) echo "curl: LDAP search failed" ;;
    44) echo "curl: Internal error (bad function call order)" ;;
    45) echo "curl: Interface error (failed to bind to specified interface)" ;;
    46) echo "curl: Bad password entered" ;;
    47) echo "curl: Too many redirects" ;;
    48) echo "curl: Unknown command line option specified" ;;
    51) echo "curl: SSL peer certificate or SSH host key verification failed" ;;
    52) echo "curl: Empty reply from server (got nothing)" ;;
    55) echo "curl: Failed sending network data" ;;
    56) echo "curl: Receive error (connection reset by peer)" ;;
    57) echo "curl: Unrecoverable poll/select error (system I/O failure)" ;;
    59) echo "curl: Couldn't use specified SSL cipher" ;;
    61) echo "curl: Bad/unrecognized transfer encoding" ;;
    63) echo "curl: Maximum file size exceeded" ;;
    75) echo "Temporary failure (retry later)" ;;
    78) echo "curl: Remote file not found (404 on FTP/file)" ;;
    79) echo "curl: SSH session error (key exchange/auth failed)" ;;
    92) echo "curl: HTTP/2 stream error (protocol violation)" ;;
    95) echo "curl: HTTP/3 layer error" ;;
    64) echo "Usage error (wrong arguments)" ;;
    65) echo "Data format error (bad input data)" ;;
    66) echo "Input file not found (cannot open input)" ;;
    67) echo "User not found (addressee unknown)" ;;
    68) echo "Host not found (hostname unknown)" ;;
    69) echo "Service unavailable" ;;
    70) echo "Internal software error" ;;
    71) echo "System error (OS-level failure)" ;;
    72) echo "Critical OS file missing" ;;
    73) echo "Cannot create output file" ;;
    74) echo "I/O error" ;;
    76) echo "Remote protocol error" ;;
    77) echo "Permission denied" ;;
    100) echo "APT: Package manager error (broken packages / dependency problems)" ;;
    101) echo "APT: Configuration error (bad sources.list, malformed config)" ;;
    102) echo "APT: Lock held by another process (dpkg/apt still running)" ;;
    103) echo "Validation: Shell is not Bash" ;;
    104) echo "Validation: Not running as root (or invoked via sudo)" ;;
    105) echo "Validation: Proxmox VE version not supported" ;;
    106) echo "Validation: Unsupported architecture (requires amd64 or arm64)" ;;
    107) echo "Validation: Kernel key parameters unreadable" ;;
    108) echo "Validation: Kernel key limits exceeded" ;;
    109) echo "Proxmox: No available container ID after max attempts" ;;
    110) echo "Proxmox: Failed to apply default.vars" ;;
    111) echo "Proxmox: App defaults file not available" ;;
    112) echo "Proxmox: Invalid install menu option" ;;
    113) echo "LXC: Under-provisioned — user aborted update" ;;
    114) echo "LXC: Storage too low — user aborted update" ;;
    115) echo "Download: install.func download failed or incomplete" ;;
    116) echo "Proxmox: Default bridge vmbr0 not found" ;;
    117) echo "LXC: Container did not reach running state" ;;
    118) echo "LXC: No IP assigned to container after timeout" ;;
    119) echo "Proxmox: No valid storage for rootdir content" ;;
    120) echo "Proxmox: No valid storage for vztmpl content" ;;
    121) echo "LXC: Container network not ready (no IP after retries)" ;;
    122) echo "LXC: No internet connectivity — user declined to continue" ;;
    123) echo "LXC: Local IP detection failed" ;;
    124) echo "Command timed out (timeout command)" ;;
    125) echo "Command failed to start (Docker daemon or execution error)" ;;
    126) echo "Command invoked cannot execute (permission problem?)" ;;
    127) echo "Command not found" ;;
    128) echo "Invalid argument to exit" ;;
    129) echo "Killed by SIGHUP (terminal closed / hangup)" ;;
    130) echo "Aborted by user (SIGINT)" ;;
    131) echo "Killed by SIGQUIT (core dumped)" ;;
    132) echo "Killed by SIGILL (illegal CPU instruction)" ;;
    134) echo "Process aborted (SIGABRT - possibly Node.js heap overflow)" ;;
    137) echo "Killed (SIGKILL / Out of memory?)" ;;
    139) echo "Segmentation fault (core dumped)" ;;
    141) echo "Broken pipe (SIGPIPE - output closed prematurely)" ;;
    143) echo "Terminated (SIGTERM)" ;;
    144) echo "Killed by signal 16 (SIGUSR1 / SIGSTKFLT)" ;;
    146) echo "Killed by signal 18 (SIGTSTP)" ;;
    150) echo "Systemd: Service failed to start" ;;
    151) echo "Systemd: Service unit not found" ;;
    152) echo "Permission denied (EACCES)" ;;
    153) echo "Build/compile failed (make/gcc/cmake)" ;;
    154) echo "Node.js: Native addon build failed (node-gyp)" ;;
    160) echo "Python: Virtualenv / uv environment missing or broken" ;;
    161) echo "Python: Dependency resolution failed" ;;
    162) echo "Python: Installation aborted (permissions or EXTERNALLY-MANAGED)" ;;
    170) echo "PostgreSQL: Connection failed (server not running / wrong socket)" ;;
    171) echo "PostgreSQL: Authentication failed (bad user/password)" ;;
    172) echo "PostgreSQL: Database does not exist" ;;
    173) echo "PostgreSQL: Fatal error in query / syntax" ;;
    180) echo "MySQL/MariaDB: Connection failed (server not running / wrong socket)" ;;
    181) echo "MySQL/MariaDB: Authentication failed (bad user/password)" ;;
    182) echo "MySQL/MariaDB: Database does not exist" ;;
    183) echo "MySQL/MariaDB: Fatal error in query / syntax" ;;
    190) echo "MongoDB: Connection failed (server not running)" ;;
    191) echo "MongoDB: Authentication failed (bad user/password)" ;;
    192) echo "MongoDB: Database not found" ;;
    193) echo "MongoDB: Fatal query error" ;;
    200) echo "Proxmox: Failed to create lock file" ;;
    203) echo "Proxmox: Missing CTID variable" ;;
    204) echo "Proxmox: Missing PCT_OSTYPE variable" ;;
    205) echo "Proxmox: Invalid CTID (<100)" ;;
    206) echo "Proxmox: CTID already in use" ;;
    207) echo "Proxmox: Password contains unescaped special characters" ;;
    208) echo "Proxmox: Invalid configuration (DNS/MAC/Network format)" ;;
    209) echo "Proxmox: Container creation failed" ;;
    210) echo "Proxmox: Cluster not quorate" ;;
    211) echo "Proxmox: Timeout waiting for template lock" ;;
    212) echo "Proxmox: Storage type 'iscsidirect' does not support containers (VMs only)" ;;
    213) echo "Proxmox: Storage type does not support 'rootdir' content" ;;
    214) echo "Proxmox: Not enough storage space" ;;
    215) echo "Proxmox: Container created but not listed (ghost state)" ;;
    216) echo "Proxmox: RootFS entry missing in config" ;;
    217) echo "Proxmox: Storage not accessible" ;;
    218) echo "Proxmox: Template file corrupted or incomplete" ;;
    219) echo "Proxmox: CephFS does not support containers - use RBD" ;;
    220) echo "Proxmox: Unable to resolve template path" ;;
    221) echo "Proxmox: Template file not readable" ;;
    222) echo "Proxmox: Template download failed" ;;
    223) echo "Proxmox: Template not available after download" ;;
    224) echo "Proxmox: PBS storage is for backups only" ;;
    225) echo "Proxmox: No template available for OS/Version" ;;
    226) echo "Proxmox: VM disk import or post-creation setup failed" ;;
    231) echo "Proxmox: LXC stack upgrade failed" ;;
    232) echo "Tools: Wrong execution environment (run on PVE host, not inside LXC)" ;;
    233) echo "Tools: Application not installed (update prerequisite missing)" ;;
    234) echo "Tools: No LXC containers found or available" ;;
    235) echo "Tools: Backup or restore operation failed" ;;
    236) echo "Tools: Required hardware not detected" ;;
    237) echo "Tools: Dependency package installation failed" ;;
    238) echo "Tools: OS or distribution not supported for this addon" ;;
    239) echo "npm/Node.js: Unexpected runtime error or dependency failure" ;;
    243) echo "Node.js: Out of memory (JavaScript heap out of memory)" ;;
    245) echo "Node.js: Invalid command-line option" ;;
    246) echo "Node.js: Internal JavaScript Parse Error" ;;
    247) echo "Node.js: Fatal internal error" ;;
    248) echo "Node.js: Invalid C++ addon / N-API failure" ;;
    249) echo "npm/pnpm/yarn: Unknown fatal error" ;;
    250) echo "App: Download failed or version not determined" ;;
    251) echo "App: File extraction failed (corrupt or incomplete archive)" ;;
    252) echo "App: Required file or resource not found" ;;
    253) echo "App: Data migration required — update aborted" ;;
    254) echo "App: User declined prompt or input timed out" ;;
    255) echo "DPKG: Fatal internal error" ;;
    *) echo "Unknown error" ;;
    esac
  }
fi

# ==============================================================================
# SECTION 2: CONTEXT DETECTION & CONTAINER ARTIFACTS
# ==============================================================================

# ------------------------------------------------------------------------------
# _is_container_context()
#
# - Returns 0 (true) when running INSIDE the LXC container being installed
# - TELEMETRY_CONTEXT can override the heuristic ("host" / "container");
#   install.func sets TELEMETRY_CONTEXT=container during bootstrap
# ------------------------------------------------------------------------------
_is_container_context() {
  case "${TELEMETRY_CONTEXT:-}" in
  container) return 0 ;;
  host) return 1 ;;
  esac
  # Proxmox/Incus tooling exists only on the host
  command -v pveversion &>/dev/null && return 1
  command -v pct &>/dev/null && return 1
  command -v incus &>/dev/null && return 1
  # systemd-detect-virt reports lxc inside containers
  if command -v systemd-detect-virt &>/dev/null; then
    case "$(systemd-detect-virt -c 2>/dev/null)" in
    lxc | lxc-libvirt | openvz) return 0 ;;
    esac
  fi
  # PCT_OSTYPE is exported into the install environment by the host
  [[ -n "${PCT_OSTYPE:-}" ]] && return 0
  return 1
}

# ------------------------------------------------------------------------------
# _container_write_failure()
#
# - Writes the failure artifacts inside the container for the host to pick up:
#   * flag file with the exit code
#   * .errinfo capture (if silent() has not already written a better one)
#   * copy of the install log
# - This REPLACES any direct telemetry send from the container
# - Arguments: $1 = exit_code, $2 = command (optional), $3 = line (optional)
# ------------------------------------------------------------------------------
_container_write_failure() {
  local exit_code="${1:-1}"
  local command="${2:-}"
  local line="${3:-}"
  local sid="${SESSION_ID:-error}"

  # Flag file with exit code (host reads this after lxc-attach returns)
  echo "$exit_code" >"/root/.install-${sid}.failed" 2>/dev/null || true

  # Keep the install log where the host expects it
  if [[ -n "${INSTALL_LOG:-}" && -f "${INSTALL_LOG}" && "${INSTALL_LOG}" != "/root/.install-${sid}.log" ]]; then
    cp "${INSTALL_LOG}" "/root/.install-${sid}.log" 2>/dev/null || true
  fi

  # Structured error capture (skip if silent() already wrote the exact
  # output segment of the failing command - that one is always better).
  # Self-contained: api.func is NOT sourced inside containers.
  local errinfo="${INSTALL_LOG:-/root/.install-${sid}.log}.errinfo"
  if [[ ! -s "$errinfo" ]]; then
    local flat_cmd
    flat_cmd=$(printf '%s' "${command:-unknown}" | tr '\n' ' ' | head -c 300)
    {
      echo "EXIT_CODE=${exit_code}"
      echo "LINE=${line:-0}"
      echo "COMMAND=${flat_cmd}"
      echo "--- OUTPUT ---"
      if [[ -n "${INSTALL_LOG:-}" && -s "${INSTALL_LOG}" ]]; then
        tail -n 80 "${INSTALL_LOG}" 2>/dev/null |
          sed 's/\r$//' |
          sed 's/\x1b\[[0-9;]*[a-zA-Z]//g' |
          grep -avE '^(Get:|Hit:|Ign:|Fetched |Reading package lists|Reading state information|Building dependency tree|Selecting previously|Preparing to unpack|Unpacking |Processing triggers for|\(Reading database|[0-9]+%[[:space:]]*\[)' |
          grep -avE '^[[:space:]]*$' |
          tail -n 60 | head -c 10240
      fi
    } >"$errinfo" 2>/dev/null || true
  fi
}

# ==============================================================================
# SECTION 3: ERROR HANDLER
# ==============================================================================

# ------------------------------------------------------------------------------
# error_handler()
#
# - Main error handler triggered by ERR trap
# - Displays error message with line number, exit code, explanation, command
# - Shows last 20 lines of the active log
# - Emits actionable hints for common failure patterns (OOM, APT, network...)
# - HOST: reports "failed" to telemetry (full payload via api.func)
# - CONTAINER: writes failure artifacts, sends nothing
# - Exits with original exit code
# ------------------------------------------------------------------------------
error_handler() {
  local exit_code=${1:-$?}
  local command=${2:-${BASH_COMMAND:-unknown}}
  local line_number=${BASH_LINENO[0]:-unknown}

  command="${command//\$STD/}"

  # If error originated from silent(), use its captured metadata
  if [[ -n "${_SILENT_FAILED_RC:-}" ]]; then
    exit_code="$_SILENT_FAILED_RC"
    command="$_SILENT_FAILED_CMD"
    line_number="$_SILENT_FAILED_LINE"
    unset _SILENT_FAILED_RC _SILENT_FAILED_CMD _SILENT_FAILED_LINE
  fi

  if [[ "$exit_code" -eq 0 ]]; then
    return 0
  fi

  # Export the failure location so telemetry can include the "where"
  FAILED_COMMAND="$command"
  FAILED_LINE="$line_number"
  export FAILED_COMMAND FAILED_LINE

  # Stop spinner and restore cursor FIRST - before any output
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h"

  local explanation
  explanation="$(explain_exit_code "$exit_code")"
  if [[ "$explanation" == curl:* && ! "$command" =~ (^|[[:space:]])([^[:space:]]*/)?curl([[:space:]]|$) ]]; then
    explanation="Command failed with exit status ${exit_code}"
  fi

  # ── Telemetry / failure artifacts ──
  if _is_container_context; then
    _container_write_failure "$exit_code" "$command" "$line_number"
  elif declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "failed" "$exit_code" 2>/dev/null || true
  fi

  # ── Display ──
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "in line ${line_number}: exit code ${exit_code} (${explanation}): while executing command ${command}"
  else
    echo -e "\n${RD}[ERROR]${CL} in line ${RD}${line_number}${CL}: exit code ${RD}${exit_code}${CL} (${explanation}): while executing command ${YWB}${command}${CL}\n"
  fi

  if [[ -n "${DEBUG_LOGFILE:-}" ]]; then
    {
      echo "------ ERROR ------"
      echo "Timestamp : $(date '+%Y-%m-%d %H:%M:%S')"
      echo "Exit Code : $exit_code ($explanation)"
      echo "Line      : $line_number"
      echo "Command   : $command"
      echo "-------------------"
    } >>"$DEBUG_LOGFILE"
  fi

  # Get active log file (prefer silent()'s logfile when available)
  local active_log=""
  if [[ -n "${_SILENT_FAILED_LOG:-}" && -s "${_SILENT_FAILED_LOG}" ]]; then
    active_log="$_SILENT_FAILED_LOG"
    unset _SILENT_FAILED_LOG
  elif declare -f get_active_logfile >/dev/null 2>&1; then
    active_log="$(get_active_logfile)"
  elif [[ -n "${SILENT_LOGFILE:-}" ]]; then
    active_log="$SILENT_LOGFILE"
  fi

  if [[ -n "$active_log" && ! -s "$active_log" && -n "${BUILD_LOG:-}" && -s "${BUILD_LOG}" ]]; then
    active_log="$BUILD_LOG"
  fi

  if [[ -n "$active_log" && -s "$active_log" ]]; then
    echo -e "\n${TAB}--- Last 20 lines of log ---"
    tail -n 20 "$active_log"
    echo -e "${TAB}-----------------------------------\n"
  fi

  # ── Node.js heap OOM detection with actionable guidance ──
  local node_oom_detected="false"
  local node_build_context="false"
  if [[ "$command" =~ (npm|pnpm|yarn|node|vite|turbo) ]]; then
    node_build_context="true"
  fi
  if [[ "$exit_code" == "243" ]]; then
    node_oom_detected="true"
  elif [[ -n "$active_log" && -s "$active_log" ]]; then
    if tail -n 200 "$active_log" 2>/dev/null | grep -Eqi 'Reached heap limit|JavaScript heap out of memory|Allocation failed - JavaScript heap out of memory|FATAL ERROR: Reached heap limit'; then
      node_oom_detected="true"
    fi
  fi

  if [[ "$node_oom_detected" == "true" ]] || { [[ "$node_build_context" == "true" ]] && [[ "$exit_code" =~ ^(134|137)$ ]]; }; then
    local heap_hint_mb=""

    if [[ -n "${NODE_OPTIONS:-}" ]] && [[ "${NODE_OPTIONS}" =~ max-old-space-size=([0-9]+) ]]; then
      heap_hint_mb="${BASH_REMATCH[1]}"
    elif [[ -n "${var_ram:-}" ]] && [[ "${var_ram}" =~ ^[0-9]+$ ]]; then
      heap_hint_mb=$((var_ram * 75 / 100))
    else
      local mem_kb=""
      mem_kb=$(awk '/^MemTotal:/ {print $2; exit}' /proc/meminfo 2>/dev/null || echo "")
      if [[ "$mem_kb" =~ ^[0-9]+$ ]]; then
        local mem_mb=$((mem_kb / 1024))
        heap_hint_mb=$((mem_mb * 75 / 100))
      fi
    fi

    if [[ -z "$heap_hint_mb" ]] || ((heap_hint_mb < 1024)); then
      heap_hint_mb=1024
    elif ((heap_hint_mb > 12288)); then
      heap_hint_mb=12288
    fi

    if declare -f msg_warn >/dev/null 2>&1; then
      msg_warn "Possible Node.js heap OOM. Try: export NODE_OPTIONS=\"--max-old-space-size=${heap_hint_mb}\" and rerun the build."
    else
      echo -e "${YW}Possible Node.js heap OOM. Try: export NODE_OPTIONS=\"--max-old-space-size=${heap_hint_mb}\" and rerun the build.${CL}"
    fi
  fi

  # ── Log-pattern analysis: actionable hints for common failure causes ──
  if [[ -n "$active_log" && -s "$active_log" ]]; then
    local _log_tail
    _log_tail=$(tail -n 60 "$active_log" 2>/dev/null || true)

    # 1. APT/dpkg dependency conflict
    if echo "$_log_tail" | grep -qE "Depends:|depends on.*but.*not installed|broken packages|unmet dep|dependency problems"; then
      local _pg_conflict
      _pg_conflict=$(echo "$_log_tail" | grep -oE 'postgresql-[0-9]+ but.*installed' | head -1 || true)
      if [[ -n "$_pg_conflict" ]]; then
        if declare -f msg_warn >/dev/null 2>&1; then
          msg_warn "PostgreSQL version conflict: ${_pg_conflict}"
          msg_warn "Hint: A package requires a specific PostgreSQL version that is not installed. Your distribution may have installed a different PG version than expected."
        fi
      else
        if declare -f msg_warn >/dev/null 2>&1; then
          msg_warn "APT dependency conflict detected. A required package is not available or is the wrong version for this system."
          msg_warn "Hint: Run 'apt-get install -f' inside the container or check that all required repositories are configured for your distribution."
        fi
      fi
    # 2. APT/GPG signature verification failure
    elif echo "$_log_tail" | grep -qE "sqv|KEYEXPIRED|NO_PUBKEY|key is not certified|signature verification failed|is not signed|Sub-process.*sqv"; then
      if declare -f msg_warn >/dev/null 2>&1; then
        msg_warn "APT repository signature error detected."
        msg_warn "Hint: A repository GPG key may be missing, expired, or the keyring file is not yet present (/usr/share/postgresql-common/pgdg/apt.postgresql.org.asc etc.)."
        msg_warn "Hint: Install the 'postgresql-common' package first, or re-add the repository with its correct signing key."
      fi
    # 3. Network / DNS failure
    elif echo "$_log_tail" | grep -qE "Could not resolve|Failed to fetch|Unable to connect|Name or service not known|Network is unreachable|curl.*resolve"; then
      if declare -f msg_warn >/dev/null 2>&1; then
        msg_warn "Network or DNS failure detected."
        msg_warn "Hint: Check the container's network connectivity, DNS settings, and whether any firewall or ad-blocker is intercepting traffic."
      fi
    # 4. APT lock held by another process
    elif echo "$_log_tail" | grep -qE "Could not get lock|dpkg frontend lock|waiting for it to exit|E: Unable to lock"; then
      if declare -f msg_warn >/dev/null 2>&1; then
        msg_warn "APT or dpkg lock conflict detected."
        msg_warn "Hint: Another package manager process may be running. Try 'rm /var/lib/dpkg/lock-frontend && dpkg --configure -a' inside the container."
      fi
    # 5. Disk space exhaustion
    elif echo "$_log_tail" | grep -qE "No space left on device|disk quota exceeded|ENOSPC"; then
      if declare -f msg_warn >/dev/null 2>&1; then
        msg_warn "Disk space exhausted during installation."
        msg_warn "Hint: Increase the container's disk size (pct resize <ctid> rootfs +2G) or clean up space first."
      fi
    fi
  fi

  # ── Context-specific cleanup ──
  if _is_container_context; then
    # Container: artifacts already written above; nothing more to do
    :
  else
    # HOST: show log path and offer container cleanup
    if [[ -n "$active_log" && -s "$active_log" ]]; then
      if declare -f msg_custom >/dev/null 2>&1; then
        msg_custom "📋" "${YW}" "Full log: ${active_log}"
      else
        echo -e "${YW}Full log:${CL} ${BL}${active_log}${CL}"
      fi
    fi

    # Offer to remove container if it exists (build errors after container creation)
    if [[ -n "${CTID:-}" ]] && command -v pct &>/dev/null && pct status "$CTID" &>/dev/null; then
      echo ""
      if declare -f msg_custom >/dev/null 2>&1; then
        echo -en "${TAB}❓${TAB}${YW}Remove broken container ${CTID}? (Y/n) [auto-remove in 60s]: ${CL}"
      else
        echo -en "${YW}Remove broken container ${CTID}? (Y/n) [auto-remove in 60s]: ${CL}"
      fi

      local response=""
      local read_rc
      read -t 60 -r response
      read_rc=$?
      if [[ $read_rc -eq 0 ]]; then
        if [[ -z "$response" || "$response" =~ ^[Yy]$ ]]; then
          echo ""
          if declare -f msg_info >/dev/null 2>&1; then
            msg_info "Removing container ${CTID}"
          else
            echo -e "${YW}Removing container ${CTID}${CL}"
          fi
          pct stop "$CTID" &>/dev/null || true
          pct destroy "$CTID" &>/dev/null || true
          if declare -f msg_ok >/dev/null 2>&1; then
            msg_ok "Container ${CTID} removed"
          else
            echo -e "${GN}✔${CL} Container ${CTID} removed"
          fi
        elif [[ "$response" =~ ^[Nn]$ ]]; then
          echo ""
          if declare -f msg_warn >/dev/null 2>&1; then
            msg_warn "Container ${CTID} kept for debugging"
          else
            echo -e "${YW}Container ${CTID} kept for debugging${CL}"
          fi
        fi
      elif [[ $read_rc -gt 128 ]]; then
        # Timeout - auto-remove
        echo ""
        if declare -f msg_info >/dev/null 2>&1; then
          msg_info "No response - removing container ${CTID}"
        else
          echo -e "${YW}No response - removing container ${CTID}${CL}"
        fi
        pct stop "$CTID" &>/dev/null || true
        pct destroy "$CTID" &>/dev/null || true
        if declare -f msg_ok >/dev/null 2>&1; then
          msg_ok "Container ${CTID} removed"
        else
          echo -e "${GN}✔${CL} Container ${CTID} removed"
        fi
      else
        # read itself failed (e.g. broken/closed stdin, I/O error) rather than
        # timing out - don't guess and destroy the container on a read we
        # couldn't actually capture; keep it since that's the reversible choice.
        echo ""
        if declare -f msg_error >/dev/null 2>&1; then
          msg_error "Could not read your response (stdin error) - keeping container ${CTID} for safety."
          msg_error "Remove it manually if not needed: pct destroy ${CTID}"
        else
          echo -e "${YW}Could not read your response (stdin error) - keeping container ${CTID} for safety.${CL}"
          echo -e "${YW}Remove it manually if not needed: pct destroy ${CTID}${CL}"
        fi
      fi
    fi
  fi

  exit "$exit_code"
}

# ==============================================================================
# SECTION 4: TELEMETRY & CLEANUP HELPERS FOR SIGNAL HANDLERS
# ==============================================================================

# ------------------------------------------------------------------------------
# _send_abort_telemetry()  (compatibility name)
#
# - HOST: reports via post_update_to_api (signals map to "aborted" there)
# - CONTAINER: writes failure artifacts instead of sending anything
# - Arguments: $1 = exit_code
# ------------------------------------------------------------------------------
_send_abort_telemetry() {
  local exit_code="${1:-1}"
  if _is_container_context; then
    _container_write_failure "$exit_code"
    return 0
  fi
  if declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "failed" "$exit_code" 2>/dev/null || true
  fi
  return 0
}

# ------------------------------------------------------------------------------
# _stop_container_if_installing()
#
# - Stops the LXC container if we're in the install phase (host only)
# - Prevents orphaned container processes when the host exits due to a signal
# ------------------------------------------------------------------------------
_stop_container_if_installing() {
  [[ "${CONTAINER_INSTALLING:-}" == "true" ]] || return 0
  [[ -n "${CTID:-}" ]] || return 0
  command -v pct &>/dev/null || return 0
  pct stop "$CTID" 2>/dev/null || true
}

# ==============================================================================
# SECTION 5: SIGNAL HANDLERS
# ==============================================================================

# ------------------------------------------------------------------------------
# on_exit()
#
# - EXIT trap handler — runs on EVERY script termination
# - CONTAINER: ensures failure artifacts exist on non-zero exit (this also
#   covers silent()'s direct `exit $rc`, which bypasses the ERR trap)
# - HOST: catches executions that never sent a final status:
#   * non-zero exit → "failed" (signal codes map to "aborted")
#   * zero exit with an "installing" record but no final → "aborted"
#     (e.g. user cancelled a whiptail dialog, script exited cleanly)
# ------------------------------------------------------------------------------
on_exit() {
  local exit_code=$?

  if _is_container_context; then
    if [[ $exit_code -ne 0 ]]; then
      _container_write_failure "$exit_code" "${FAILED_COMMAND:-}" "${FAILED_LINE:-}"
    fi
  else
    if [[ "${POST_UPDATE_DONE:-}" != "true" ]] && declare -f post_update_to_api >/dev/null 2>&1; then
      if [[ $exit_code -ne 0 ]]; then
        post_update_to_api "failed" "$exit_code" 2>/dev/null || true
      elif [[ "${POST_TO_API_DONE:-}" == "true" ]]; then
        # Clean exit but no success was ever reported: the user backed out
        # somewhere. Report as aborted so the record doesn't stay "installing".
        post_update_to_api "aborted" "0" 2>/dev/null || true
      fi
    fi

    # Best-effort log collection on failure
    if [[ $exit_code -ne 0 ]] && declare -f ensure_log_on_host >/dev/null 2>&1; then
      ensure_log_on_host 2>/dev/null || true
    fi

    # Stop orphaned container if we're in the install phase
    if [[ $exit_code -ne 0 ]]; then
      _stop_container_if_installing
    fi
  fi

  [[ -n "${lockfile:-}" && -e "$lockfile" ]] && rm -f "$lockfile"
  exit "$exit_code"
}

# ------------------------------------------------------------------------------
# on_interrupt() - SIGINT (Ctrl+C)
# ------------------------------------------------------------------------------
on_interrupt() {
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h" 2>/dev/null || true

  if _is_container_context; then
    _container_write_failure "130"
  elif declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "aborted" "130" 2>/dev/null || true
  fi
  _stop_container_if_installing
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "Interrupted by user (SIGINT)" 2>/dev/null || true
  else
    echo -e "\n${RD}Interrupted by user (SIGINT)${CL}" 2>/dev/null || true
  fi
  exit 130
}

# ------------------------------------------------------------------------------
# on_terminate() - SIGTERM
# ------------------------------------------------------------------------------
on_terminate() {
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h" 2>/dev/null || true

  if _is_container_context; then
    _container_write_failure "143"
  elif declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "aborted" "143" 2>/dev/null || true
  fi
  _stop_container_if_installing
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "Terminated by signal (SIGTERM)" 2>/dev/null || true
  else
    echo -e "\n${RD}Terminated by signal (SIGTERM)${CL}" 2>/dev/null || true
  fi
  exit 143
}

# ------------------------------------------------------------------------------
# on_hangup() - SIGHUP (SSH disconnect, terminal closed)
# ------------------------------------------------------------------------------
on_hangup() {
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi

  if _is_container_context; then
    _container_write_failure "129"
  elif declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "aborted" "129" 2>/dev/null || true
  fi
  _stop_container_if_installing
  exit 129
}

# ==============================================================================
# SECTION 6: INITIALIZATION
# ==============================================================================

# ------------------------------------------------------------------------------
# catch_errors()
#
# - Initializes error handling and signal traps
# - set -Ee -o pipefail (+ set -u when STRICT_UNSET=1)
# - Traps: ERR → error_handler, EXIT → on_exit, INT/TERM/HUP → signal handlers
# - Call this function early in every script
# ------------------------------------------------------------------------------
catch_errors() {
  set -Ee -o pipefail
  if [ "${STRICT_UNSET:-0}" = "1" ]; then
    set -u
  fi

  trap 'error_handler' ERR
  trap on_exit EXIT
  trap on_interrupt INT
  trap on_terminate TERM
  trap on_hangup HUP
}