community-scripts/ProxmoxVED
1
0
Fork 0
mirror of https://github.com/community-scripts/ProxmoxVED.git synced 2026-02-24 21:47:26 +00:00
Code Issues Packages Projects Releases Wiki Activity

merge build.func

Browse Source
This commit is contained in:
CanbiZ (MickLesk)
2026-02-24 09:47:12 +01:00
parent 4e2b64524f
commit d16008181d
1 changed files with 467 additions and 62 deletions
Download Patch File Download Diff File Expand all files Collapse all files

529
misc/build.func
View File

@@ -45,6 +45,7 @@ variables() {
DIAGNOSTICS="yes" # sets the DIAGNOSTICS variable to "yes", used for the API call.
METHOD="default" # sets the METHOD variable to "default", used for the API call.
RANDOM_UUID="$(cat /proc/sys/kernel/random/uuid)" # generates a random UUID and sets it to the RANDOM_UUID variable.
SESSION_ID="${RANDOM_UUID:0:8}" # Short session ID (first 8 chars of UUID) for log files
BUILD_LOG="/tmp/create-lxc-${SESSION_ID}.log" # Host-side container creation log
combined_log="/tmp/install-${SESSION_ID}-combined.log" # Combined log (build + install) for failed installations
@@ -58,7 +59,7 @@ variables() {
mkdir -p /var/log/community-scripts
BUILD_LOG="/var/log/community-scripts/create-lxc-${SESSION_ID}-$(date +%Y%m%d_%H%M%S).log"
combined_log="/var/log/community-scripts/install-${SESSION_ID}-combined-$(date +%Y%m%d_%H%M%S).log"
fi
fi
# Get Proxmox VE version and kernel version
if command -v pveversion >/dev/null 2>&1; then
@@ -213,6 +214,26 @@ update_motd_ip() {
# Add the new IP address
echo -e "${TAB}${NETWORK}${YW} IP Address: ${GN}${IP}${CL}" >>"$MOTD_FILE"
fi
# Update dynamic LXC details profile if values changed (e.g., after OS upgrade)
# Only update if file exists and is from community-scripts
if [ -f "$PROFILE_FILE" ] && grep -q "community-scripts" "$PROFILE_FILE" 2>/dev/null; then
# Get current values
local current_os="$(grep ^NAME /etc/os-release | cut -d= -f2 | tr -d '"') - Version: $(grep ^VERSION_ID /etc/os-release | cut -d= -f2 | tr -d '"')"
local current_hostname="$(hostname)"
local current_ip="$(hostname -I | awk '{print $1}')"
# Update only if values actually changed
if ! grep -q "OS:.*$current_os" "$PROFILE_FILE" 2>/dev/null; then
sed -i "s|OS:.*|OS: \${GN}$current_os\${CL}\\\"|" "$PROFILE_FILE"
fi
if ! grep -q "Hostname:.*$current_hostname" "$PROFILE_FILE" 2>/dev/null; then
sed -i "s|Hostname:.*|Hostname: \${GN}$current_hostname\${CL}\\\"|" "$PROFILE_FILE"
fi
if ! grep -q "IP Address:.*$current_ip" "$PROFILE_FILE" 2>/dev/null; then
sed -i "s|IP Address:.*|IP Address: \${GN}$current_ip\${CL}\\\"|" "$PROFILE_FILE"
fi
fi
}
# ------------------------------------------------------------------------------
@@ -3391,6 +3412,69 @@ configure_ssh_settings() {
fi
}
# ------------------------------------------------------------------------------
# msg_menu()
#
# - Displays a numbered menu for update_script() functions
# - In silent mode (PHS_SILENT=1): auto-selects the default option
# - In interactive mode: shows menu via read with 10s timeout + default fallback
# - Usage: CHOICE=$(msg_menu "Title" "tag1" "Description 1" "tag2" "Desc 2" ...)
# - The first item is always the default
# - Returns the selected tag to stdout
# - If no valid selection or timeout, returns the default (first) tag
# ------------------------------------------------------------------------------
msg_menu() {
local title="$1"
shift
# Parse items into parallel arrays: tags[] and descriptions[]
local -a tags=()
local -a descs=()
while [[ $# -ge 2 ]]; do
tags+=("$1")
descs+=("$2")
shift 2
done
local default_tag="${tags[0]}"
local count=${#tags[@]}
# Silent mode: return default immediately
if [[ -n "${PHS_SILENT+x}" ]] && [[ "${PHS_SILENT}" == "1" ]]; then
echo "$default_tag"
return 0
fi
# Display menu to /dev/tty so it doesn't get captured by command substitution
{
echo ""
msg_custom "📋" "${BL}" "${title}"
echo ""
for i in "${!tags[@]}"; do
local marker=" "
[[ $i -eq 0 ]] && marker="* "
printf "${TAB3}${marker}%s) %s\n" "${tags[$i]}" "${descs[$i]}"
done
echo ""
} >/dev/tty
local selection=""
read -r -t 10 -p "${TAB3}Select [default=${default_tag}, timeout 10s]: " selection </dev/tty >/dev/tty || true
# Validate selection
if [[ -n "$selection" ]]; then
for tag in "${tags[@]}"; do
if [[ "$selection" == "$tag" ]]; then
echo "$selection"
return 0
fi
done
msg_warn "Invalid selection '${selection}' - using default: ${default_tag}"
fi
echo "$default_tag"
return 0
}
# ------------------------------------------------------------------------------
# start()
#
@@ -3538,6 +3622,7 @@ build_container() {
# Core exports for install.func
export DIAGNOSTICS="$DIAGNOSTICS"
export RANDOM_UUID="$RANDOM_UUID"
export EXECUTION_ID="$EXECUTION_ID"
export SESSION_ID="$SESSION_ID"
export CACHER="$APT_CACHER"
export CACHER_IP="$APT_CACHER_IP"
@@ -3563,6 +3648,11 @@ build_container() {
export BUILD_LOG="$BUILD_LOG"
export INSTALL_LOG="/root/.install-${SESSION_ID}.log"
export COMMUNITY_SCRIPTS_URL="$COMMUNITY_SCRIPTS_URL"
# Keep host-side logging on BUILD_LOG (not exported — invisible to container)
# Without this, get_active_logfile() would return INSTALL_LOG (a container path)
# and all host msg_info/msg_ok/msg_error would write to /root/.install-SESSION.log
# on the HOST instead of BUILD_LOG, causing incomplete telemetry logs.
_HOST_LOGFILE="$BUILD_LOG"
export dev_mode="${dev_mode:-}"
export DEV_MODE_MOTD="${DEV_MODE_MOTD:-false}"
export DEV_MODE_KEEP="${DEV_MODE_KEEP:-false}"
@@ -3649,13 +3739,11 @@ $PCT_OPTIONS_STRING"
exit 214
fi
msg_ok "Storage space validated"
# Report installation start to API (early - captures failed installs too)
post_to_api
fi
create_lxc_container || exit $?
# Transition to 'configuring' — container created, now setting up OS/userland
post_progress_to_api "configuring"
LXC_CONFIG="/etc/pve/lxc/${CTID}.conf"
# ============================================================================
@@ -4130,11 +4218,14 @@ EOF'
exit $install_exit_code
fi
# Prompt user for cleanup with 60s timeout (plain echo - no msg_info to avoid spinner)
# Prompt user for cleanup with 60s timeout
echo ""
# Detect error type for smart recovery options
local is_oom=false
local is_network_issue=false
local is_apt_issue=false
local is_cmd_not_found=false
local error_explanation=""
if declare -f explain_exit_code >/dev/null 2>&1; then
error_explanation="$(explain_exit_code "$install_exit_code")"
@@ -4145,26 +4236,127 @@ EOF'
is_oom=true
fi
# APT/DPKG detection: exit codes 100-102 (APT), 255 (DPKG with log evidence)
case "$install_exit_code" in
100 | 101 | 102) is_apt_issue=true ;;
255)
if [[ -f "$combined_log" ]] && grep -qiE 'dpkg|apt-get|apt\.conf|broken packages|unmet dependencies|E: Sub-process|E: Failed' "$combined_log"; then
is_apt_issue=true
fi
;;
esac
# Command not found detection
if [[ $install_exit_code -eq 127 ]]; then
is_cmd_not_found=true
fi
# Network-related detection (curl/apt/git fetch failures and transient network issues)
case "$install_exit_code" in
6 | 7 | 22 | 28 | 35 | 52 | 56 | 57 | 75 | 78) is_network_issue=true ;;
100)
# APT can fail due to network (Failed to fetch)
if [[ -f "$combined_log" ]] && grep -qiE 'Failed to fetch|Could not resolve|Connection failed|Network is unreachable|Temporary failure resolving' "$combined_log"; then
is_network_issue=true
fi
;;
128)
if [[ -f "$combined_log" ]] && grep -qiE 'RPC failed|early EOF|fetch-pack|HTTP/2 stream|Could not resolve host|Temporary failure resolving|Failed to fetch|Connection reset|Network is unreachable' "$combined_log"; then
is_network_issue=true
fi
;;
esac
# Exit 1 subclassification: analyze logs to identify actual root cause
# Many exit 1 errors are actually APT, OOM, network, or command-not-found issues
if [[ $install_exit_code -eq 1 && -f "$combined_log" ]]; then
if grep -qiE 'E: Unable to|E: Package|E: Failed to fetch|dpkg.*error|broken packages|unmet dependencies|dpkg --configure -a' "$combined_log"; then
is_apt_issue=true
fi
if grep -qiE 'Cannot allocate memory|Out of memory|oom-killer|Killed process|JavaScript heap' "$combined_log"; then
is_oom=true
fi
if grep -qiE 'Could not resolve|DNS|Connection refused|Network is unreachable|No route to host|Temporary failure resolving|Failed to fetch' "$combined_log"; then
is_network_issue=true
fi
if grep -qiE ': command not found|No such file or directory.*/s?bin/' "$combined_log"; then
is_cmd_not_found=true
fi
fi
# Show error explanation if available
if [[ -n "$error_explanation" ]]; then
echo -e "${TAB}${RD}Error: ${error_explanation}${CL}"
echo ""
fi
# Show specific hints for known error types
if [[ $install_exit_code -eq 10 ]]; then
echo -e "${TAB}${INFO} This error usually means the container needs ${GN}privileged${CL} mode or Docker/nesting support."
echo -e "${TAB}${INFO} Recreate with: Advanced Install → Container Type: ${GN}Privileged${CL}"
echo ""
fi
if [[ $install_exit_code -eq 125 || $install_exit_code -eq 126 ]]; then
echo -e "${TAB}${INFO} The command exists but cannot be executed. This may be a ${GN}permission${CL} issue."
echo -e "${TAB}${INFO} If using Docker, ensure the container is ${GN}privileged${CL} or has correct permissions."
echo ""
fi
if [[ "$is_cmd_not_found" == true ]]; then
local missing_cmd=""
if [[ -f "$combined_log" ]]; then
missing_cmd=$(grep -oiE '[a-zA-Z0-9_.-]+: command not found' "$combined_log" | tail -1 | sed 's/: command not found//')
fi
if [[ -n "$missing_cmd" ]]; then
echo -e "${TAB}${INFO} Missing command: ${GN}${missing_cmd}${CL}"
fi
echo ""
fi
# Build recovery menu based on error type
echo -e "${YW}What would you like to do?${CL}"
echo ""
echo -e " ${GN}1)${CL} Remove container and exit"
echo -e " ${GN}2)${CL} Keep container for debugging"
echo -e " ${GN}3)${CL} Retry with verbose mode"
if [[ "$is_oom" == true ]]; then
local new_ram=$((RAM_SIZE * 3 / 2))
local new_cpu=$((CORE_COUNT + 1))
echo -e " ${GN}4)${CL} Retry with more resources (RAM: ${RAM_SIZE}→${new_ram} MiB, CPU: ${CORE_COUNT}→${new_cpu} cores)"
fi
echo ""
echo -en "${YW}Select option [1-$([[ "$is_oom" == true ]] && echo "4" || echo "3")] (default: 1, auto-remove in 60s): ${CL}"
echo -e " ${GN}3)${CL} Retry with verbose mode (full rebuild)"
local next_option=4
local APT_OPTION="" OOM_OPTION="" DNS_OPTION=""
if [[ "$is_apt_issue" == true ]]; then
if [[ "$var_os" == "alpine" ]]; then
echo -e " ${GN}${next_option})${CL} Repair APK state and re-run install (in-place)"
else
echo -e " ${GN}${next_option})${CL} Repair APT/DPKG state and re-run install (in-place)"
fi
APT_OPTION=$next_option
next_option=$((next_option + 1))
fi
if [[ "$is_oom" == true ]]; then
local recovery_attempt="${RECOVERY_ATTEMPT:-0}"
if [[ $recovery_attempt -lt 2 ]]; then
local new_ram=$((RAM_SIZE * 2))
local new_cpu=$((CORE_COUNT * 2))
echo -e " ${GN}${next_option})${CL} Retry with more resources (RAM: ${RAM_SIZE}→${new_ram} MiB, CPU: ${CORE_COUNT}→${new_cpu} cores)"
OOM_OPTION=$next_option
next_option=$((next_option + 1))
else
echo -e " ${DGN}-)${CL} ${DGN}OOM retry exhausted (already retried ${recovery_attempt}x)${CL}"
fi
fi
if [[ "$is_network_issue" == true ]]; then
echo -e " ${GN}${next_option})${CL} Retry with DNS override in LXC (8.8.8.8 / 1.1.1.1)"
DNS_OPTION=$next_option
next_option=$((next_option + 1))
fi
local max_option=$((next_option - 1))
echo ""
echo -en "${YW}Select option [1-${max_option}] (default: 1, auto-remove in 60s): ${CL}"
if read -t 60 -r response; then
case "${response:-1}" in
1)
@@ -4202,6 +4394,7 @@ EOF'
export VERBOSE="yes"
export var_verbose="yes"
# Show rebuild summary
echo -e "${YW}Rebuilding with preserved settings:${CL}"
echo -e " Container ID: ${old_ctid} → ${CTID}"
@@ -4214,56 +4407,166 @@ EOF'
build_container
return $?
;;
4)
if [[ "$is_oom" == true ]]; then
# Retry with more resources
*)
# Handle dynamic smart recovery options via named option variables
local handled=false
if [[ -n "${APT_OPTION}" && "${response}" == "${APT_OPTION}" ]]; then
# Package manager in-place repair: fix broken state and re-run install script
handled=true
if [[ "$var_os" == "alpine" ]]; then
echo -e "\n${TAB}${HOLD}${YW}Repairing APK state in container ${CTID}...${CL}"
pct exec "$CTID" -- ash -c "
apk fix 2>/dev/null || true
apk cache clean 2>/dev/null || true
apk update 2>/dev/null || true
" >/dev/null 2>&1 || true
echo -e "${BFR}${CM}${GN}APK state repaired in container ${CTID}${CL}"
else
echo -e "\n${TAB}${HOLD}${YW}Repairing APT/DPKG state in container ${CTID}...${CL}"
pct exec "$CTID" -- bash -c "
DEBIAN_FRONTEND=noninteractive dpkg --configure -a 2>/dev/null || true
apt-get -f install -y 2>/dev/null || true
apt-get clean 2>/dev/null
apt-get update 2>/dev/null || true
" >/dev/null 2>&1 || true
echo -e "${BFR}${CM}${GN}APT/DPKG state repaired in container ${CTID}${CL}"
fi
echo ""
export VERBOSE="yes"
export var_verbose="yes"
echo -e "${YW}Re-running installation in existing container ${CTID}:${CL}"
echo -e " RAM: ${RAM_SIZE} MiB | CPU: ${CORE_COUNT} cores | Disk: ${DISK_SIZE} GB"
echo -e " Verbose: ${GN}enabled${CL}"
echo ""
msg_info "Re-running installation script..."
# Re-run install script in existing container (don't destroy/recreate)
set +Eeuo pipefail
trap - ERR
local _LXC_CAPTURE_LOG="/tmp/.install-capture-${SESSION_ID}.log"
lxc-attach -n "$CTID" -- bash -c "$(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/install/${var_install}.sh)" 2>&1 | tee "$_LXC_CAPTURE_LOG"
local apt_retry_exit=${PIPESTATUS[0]}
set -Eeuo pipefail
trap 'error_handler' ERR
# Check for error flag from retry
local apt_retry_code=0
if [[ -n "${SESSION_ID:-}" ]]; then
local retry_error_flag="/root/.install-${SESSION_ID}.failed"
if pct exec "$CTID" -- test -f "$retry_error_flag" 2>/dev/null; then
apt_retry_code=$(pct exec "$CTID" -- cat "$retry_error_flag" 2>/dev/null || echo "1")
pct exec "$CTID" -- rm -f "$retry_error_flag" 2>/dev/null || true
fi
fi
if [[ $apt_retry_code -eq 0 && $apt_retry_exit -ne 0 ]]; then
apt_retry_code=$apt_retry_exit
fi
if [[ $apt_retry_code -eq 0 ]]; then
msg_ok "Installation completed successfully after APT repair!"
post_update_to_api "done" "0" "force"
return 0
else
msg_error "Installation still failed after APT repair (exit code: ${apt_retry_code})"
install_exit_code=$apt_retry_code
fi
fi
if [[ -n "${OOM_OPTION}" && "${response}" == "${OOM_OPTION}" ]]; then
# Retry with doubled resources
handled=true
echo -e "\n${TAB}${HOLD}${YW}Removing container ${CTID} for rebuild with more resources...${CL}"
pct stop "$CTID" &>/dev/null || true
pct destroy "$CTID" &>/dev/null || true
echo -e "${BFR}${CM}${GN}Container ${CTID} removed${CL}"
echo ""
# Get new container ID and increase resources
local old_ctid="$CTID"
local old_ram="$RAM_SIZE"
local old_cpu="$CORE_COUNT"
export CTID=$(get_valid_container_id "$CTID")
export RAM_SIZE=$((RAM_SIZE * 3 / 2))
export CORE_COUNT=$((CORE_COUNT + 1))
export RAM_SIZE=$((RAM_SIZE * 2))
export CORE_COUNT=$((CORE_COUNT * 2))
export var_ram="$RAM_SIZE"
export var_cpu="$CORE_COUNT"
export VERBOSE="yes"
export var_verbose="yes"
export RECOVERY_ATTEMPT=$((${RECOVERY_ATTEMPT:-0} + 1))
# Show rebuild summary
echo -e "${YW}Rebuilding with increased resources:${CL}"
echo -e "${YW}Rebuilding with increased resources (attempt ${RECOVERY_ATTEMPT}/2):${CL}"
echo -e " Container ID: ${old_ctid} → ${CTID}"
echo -e " RAM: ${old_ram} → ${GN}${RAM_SIZE}${CL} MiB (+50%)"
echo -e " CPU: ${old_cpu} → ${GN}${CORE_COUNT}${CL} cores (+1)"
echo -e " RAM: ${old_ram} → ${GN}${RAM_SIZE}${CL} MiB (x2)"
echo -e " CPU: ${old_cpu} → ${GN}${CORE_COUNT}${CL} cores (x2)"
echo -e " Disk: ${DISK_SIZE} GB | Network: ${NET:-dhcp} | Bridge: ${BRG:-vmbr0}"
echo -e " Verbose: ${GN}enabled${CL}"
echo ""
msg_info "Restarting installation..."
# Re-run build_container
build_container
return $?
else
fi
if [[ -n "${DNS_OPTION}" && "${response}" == "${DNS_OPTION}" ]]; then
# Retry with DNS override in LXC
handled=true
echo -e "\n${TAB}${HOLD}${YW}Removing container ${CTID} for rebuild with DNS override...${CL}"
pct stop "$CTID" &>/dev/null || true
pct destroy "$CTID" &>/dev/null || true
echo -e "${BFR}${CM}${GN}Container ${CTID} removed${CL}"
echo ""
local old_ctid="$CTID"
export CTID=$(get_valid_container_id "$CTID")
export DNS_RETRY_OVERRIDE="true"
export VERBOSE="yes"
export var_verbose="yes"
echo -e "${YW}Rebuilding with DNS override in LXC:${CL}"
echo -e " Container ID: ${old_ctid} → ${CTID}"
echo -e " DNS: ${GN}8.8.8.8, 1.1.1.1${CL} (inside LXC only)"
echo -e " Verbose: ${GN}enabled${CL}"
echo ""
msg_info "Restarting installation..."
build_container
return $?
fi
if [[ "$handled" == false ]]; then
echo -e "\n${TAB}${YW}Invalid option. Container ${CTID} kept.${CL}"
exit $install_exit_code
fi
;;
*)
echo -e "\n${TAB}${YW}Invalid option. Container ${CTID} kept.${CL}"
exit $install_exit_code
;;
esac
else
# Timeout - auto-remove
echo -e "\n${YW}No response - auto-removing container${CL}"
echo -e "${TAB}${HOLD}${YW}Removing container ${CTID}${CL}"
echo ""
msg_info "No response - removing container ${CTID}"
pct stop "$CTID" &>/dev/null || true
pct destroy "$CTID" &>/dev/null || true
echo -e "${BFR}${CM}${GN}Container ${CTID} removed${CL}"
msg_ok "Container ${CTID} removed"
fi
# Force one final status update attempt after cleanup
# This ensures status is updated even if the first attempt failed (e.g., HTTP 400)
post_update_to_api "failed" "$install_exit_code" "force"
exit $install_exit_code
fi
# Clean up host-side capture log (not needed on success, already in combined_log on failure)
rm -f "/tmp/.install-capture-${SESSION_ID}.log" 2>/dev/null
# Re-enable error handling after successful install or recovery menu completion
set -Eeuo pipefail
trap 'error_handler' ERR
}
destroy_lxc() {
@@ -4275,16 +4578,29 @@ destroy_lxc() {
# Abort on Ctrl-C / Ctrl-D / ESC
trap 'echo; msg_error "Aborted by user (SIGINT/SIGQUIT)"; return 130' INT QUIT
if prompt_confirm "Remove this Container?" "n" 60; then
local prompt
if ! read -rp "Remove this Container? <y/N> " prompt; then
# read returns non-zero on Ctrl-D/ESC
msg_error "Aborted input (Ctrl-D/ESC)"
return 130
fi
case "${prompt,,}" in
y | yes)
if pct stop "$CT_ID" &>/dev/null && pct destroy "$CT_ID" &>/dev/null; then
msg_ok "Removed Container $CT_ID"
else
msg_error "Failed to remove Container $CT_ID"
return 1
fi
else
;;
"" | n | no)
msg_custom "ℹ️" "${BL}" "Container was not removed."
fi
;;
*)
msg_warn "Invalid response. Container was not removed."
;;
esac
}
# ------------------------------------------------------------------------------
@@ -4653,6 +4969,12 @@ create_lxc_container() {
exit 206
fi
# Report installation start to API early - captures failures in storage/template/create
post_to_api
# Transition to 'validation' — Proxmox-internal checks (storage, template, cluster)
post_progress_to_api "validation"
# Storage capability check
check_storage_support "rootdir" || {
msg_error "No valid storage found for 'rootdir' [Container]"
@@ -4797,29 +5119,37 @@ create_lxc_container() {
)
if [[ ${#AVAILABLE_VERSIONS[@]} -gt 0 ]]; then
# Use prompt_select for version selection (supports unattended mode)
local selected_version
selected_version=$(prompt_select "Select ${PCT_OSTYPE} version:" 1 60 "${AVAILABLE_VERSIONS[@]}")
echo ""
echo "${BL}Available ${PCT_OSTYPE} versions:${CL}"
for i in "${!AVAILABLE_VERSIONS[@]}"; do
echo " [$((i + 1))] ${AVAILABLE_VERSIONS[$i]}"
done
echo ""
read -p "Select version [1-${#AVAILABLE_VERSIONS[@]}] or press Enter to cancel: " choice
# prompt_select always returns a value (uses default in unattended mode)
PCT_OSVERSION="$selected_version"
TEMPLATE_SEARCH="${PCT_OSTYPE}-${PCT_OSVERSION}"
if [[ "$choice" =~ ^[0-9]+$ ]] && [[ "$choice" -ge 1 ]] && [[ "$choice" -le ${#AVAILABLE_VERSIONS[@]} ]]; then
PCT_OSVERSION="${AVAILABLE_VERSIONS[$((choice - 1))]}"
TEMPLATE_SEARCH="${PCT_OSTYPE}-${PCT_OSVERSION}"
ONLINE_TEMPLATES=()
mapfile -t ONLINE_TEMPLATES < <(
pveam available -section system 2>/dev/null |
grep -E '\.(tar\.zst|tar\.xz|tar\.gz)$' |
awk '{print $2}' |
grep -E "^${TEMPLATE_SEARCH}-.*${TEMPLATE_PATTERN}" |
sort -t - -k 2 -V 2>/dev/null || true
)
ONLINE_TEMPLATES=()
mapfile -t ONLINE_TEMPLATES < <(
pveam available -section system 2>/dev/null |
grep -E '\.(tar\.zst|tar\.xz|tar\.gz)$' |
awk '{print $2}' |
grep -E "^${TEMPLATE_SEARCH}-.*${TEMPLATE_PATTERN}" |
sort -t - -k 2 -V 2>/dev/null || true
)
if [[ ${#ONLINE_TEMPLATES[@]} -gt 0 ]]; then
TEMPLATE="${ONLINE_TEMPLATES[-1]}"
TEMPLATE_SOURCE="online"
if [[ ${#ONLINE_TEMPLATES[@]} -gt 0 ]]; then
TEMPLATE="${ONLINE_TEMPLATES[-1]}"
TEMPLATE_SOURCE="online"
else
msg_error "No templates available for ${PCT_OSTYPE} ${PCT_OSVERSION}"
exit 225
fi
else
msg_error "No templates available for ${PCT_OSTYPE} ${PCT_OSVERSION}"
exit 225
msg_custom "🚫" "${YW}" "Installation cancelled"
exit 0
fi
else
msg_error "No ${PCT_OSTYPE} templates available at all"
@@ -5231,25 +5561,100 @@ EOF
# SECTION 10: ERROR HANDLING & EXIT TRAPS
# ==============================================================================
# ------------------------------------------------------------------------------
# ensure_log_on_host()
#
# - Ensures INSTALL_LOG points to a readable file on the host
# - If INSTALL_LOG points to a container path (e.g. /root/.install-*),
# tries to pull it from the container and create a combined log
# - This allows get_error_text() to find actual error output for telemetry
# - Uses timeout on pct pull to prevent hangs on dead/unresponsive containers
# ------------------------------------------------------------------------------
ensure_log_on_host() {
# Already readable on host? Nothing to do.
[[ -n "${INSTALL_LOG:-}" && -s "${INSTALL_LOG}" ]] && return 0
# Try pulling from container and creating combined log
if [[ -n "${CTID:-}" && -n "${SESSION_ID:-}" ]] && command -v pct &>/dev/null; then
local combined_log="/tmp/${NSAPP:-lxc}-${CTID}-${SESSION_ID}.log"
if [[ ! -s "$combined_log" ]]; then
# Create combined log
{
echo "================================================================================"
echo "COMBINED INSTALLATION LOG - ${APP:-LXC}"
echo "Container ID: ${CTID}"
echo "Session ID: ${SESSION_ID}"
echo "Timestamp: $(date '+%Y-%m-%d %H:%M:%S')"
echo "================================================================================"
echo ""
} >"$combined_log" 2>/dev/null || return 0
# Append BUILD_LOG if it exists
if [[ -f "${BUILD_LOG:-}" ]]; then
{
echo "================================================================================"
echo "PHASE 1: CONTAINER CREATION (Host)"
echo "================================================================================"
cat "${BUILD_LOG}"
echo ""
} >>"$combined_log"
fi
# Pull INSTALL_LOG from container (with timeout to prevent hangs on dead containers)
local temp_log="/tmp/.install-temp-${SESSION_ID}.log"
if timeout 8 pct pull "$CTID" "/root/.install-${SESSION_ID}.log" "$temp_log" 2>/dev/null; then
{
echo "================================================================================"
echo "PHASE 2: APPLICATION INSTALLATION (Container)"
echo "================================================================================"
cat "$temp_log"
echo ""
} >>"$combined_log"
rm -f "$temp_log"
fi
fi
if [[ -s "$combined_log" ]]; then
INSTALL_LOG="$combined_log"
fi
fi
}
# ------------------------------------------------------------------------------
# api_exit_script()
#
# - Exit trap handler for reporting to API telemetry
# - Captures exit code and reports to PocketBase using centralized error descriptions
# - Uses explain_exit_code() from api.func for consistent error messages
# - Posts failure status with exit code to API (error description resolved automatically)
# - Only executes on non-zero exit codes
# - ALWAYS sends telemetry FIRST before log collection to prevent pct pull
# hangs from blocking status updates (container may be dead/unresponsive)
# - For non-zero exit codes: posts "failed" status
# - For zero exit codes where post_update_to_api was never called:
# catches orphaned "installing" records (e.g., script exited cleanly
# but description() was never reached)
# ------------------------------------------------------------------------------
api_exit_script() {
exit_code=$?
local exit_code=$?
if [ $exit_code -ne 0 ]; then
post_update_to_api "failed" "$exit_code"
# ALWAYS send telemetry FIRST - ensure status is reported even if
# ensure_log_on_host hangs (e.g. pct pull on dead container)
post_update_to_api "failed" "$exit_code" 2>/dev/null || true
# Best-effort log collection (non-critical after telemetry is sent)
if declare -f ensure_log_on_host >/dev/null 2>&1; then
ensure_log_on_host 2>/dev/null || true
fi
# Stop orphaned container if we're in the install phase
if [[ "${CONTAINER_INSTALLING:-}" == "true" && -n "${CTID:-}" ]] && command -v pct &>/dev/null; then
pct stop "$CTID" 2>/dev/null || true
fi
elif [[ "${POST_TO_API_DONE:-}" == "true" && "${POST_UPDATE_DONE:-}" != "true" ]]; then
# Script exited with 0 but never sent a completion status
# exit_code=0 is never an error — report as success
post_update_to_api "done" "0"
fi
}
if command -v pveversion >/dev/null 2>&1; then
trap 'api_exit_script' EXIT
fi
trap 'post_update_to_api "failed" "$?"' ERR
trap 'post_update_to_api "failed" "130"' SIGINT
trap 'post_update_to_api "failed" "143"' SIGTERM
trap 'local _ec=$?; if [[ $_ec -ne 0 ]]; then post_update_to_api "failed" "$_ec" 2>/dev/null || true; if declare -f ensure_log_on_host &>/dev/null; then ensure_log_on_host 2>/dev/null || true; fi; fi' ERR
trap 'post_update_to_api "failed" "129" 2>/dev/null || true; if [[ -n "${CTID:-}" ]] && command -v pct &>/dev/null; then pct stop "$CTID" 2>/dev/null || true; fi; exit 129' SIGHUP
trap 'post_update_to_api "failed" "130" 2>/dev/null || true; if [[ -n "${CTID:-}" ]] && command -v pct &>/dev/null; then pct stop "$CTID" 2>/dev/null || true; fi; exit 130' SIGINT
trap 'post_update_to_api "failed" "143" 2>/dev/null || true; if [[ -n "${CTID:-}" ]] && command -v pct &>/dev/null; then pct stop "$CTID" 2>/dev/null || true; fi; exit 143' SIGTERM