Files
ProxmoxVE/tools/pve/fstrim.sh

194 lines
6.5 KiB
Bash
Raw Normal View History

2024-02-15 23:04:55 -05:00
#!/usr/bin/env bash
set -eEuo pipefail
2024-02-15 23:04:55 -05:00
function header_info() {
clear
cat <<"EOF"
_______ __ __ ______ _
/ ____(_) /__ _______ _______/ /____ ____ ___ /_ __/____(_)___ ___
/ /_ / / / _ \/ ___/ / / / ___/ __/ _ \/ __ `__ \ / / / ___/ / __ `__ \
/ __/ / / / __(__ ) /_/ (__ ) /_/ __/ / / / / / / / / / / / / / / / /
/_/ /_/_/\___/____/\__, /____/\__/\___/_/ /_/ /_/ /_/ /_/ /_/_/ /_/ /_/
/____/
EOF
}
BL="\033[36m"
RD="\033[01;31m"
GN="\033[1;92m"
CL="\033[m"
tools/pve: add data analytics / formatting / linting (#12034) * core: add progress; fix exit status Introduce post_progress_to_api() in alpine-install.func and install.func to send a lightweight, fire-and-forget telemetry ping (HTTP POST) that updates an existing telemetry record to "configuring" when DIAGNOSTICS=yes and RANDOM_UUID is set. The function is non-blocking (curl -m 5, errors ignored) and is invoked during container setup and after OS updates to signal active progress. Also adjust api_exit_script() in build.func to report success (post_update_to_api "done" "0") for cases where the script exited normally but a completion status wasn't posted, instead of reporting failure. * Safer tools.func load and improved error handling Replace process-substitution sourcing of tools.func with an explicit curl -> variable -> source via /dev/stdin, adding failure messages and a check that expected functions (e.g. fetch_and_deploy_gh_release) are present (misc/alpine-install.func, misc/install.func). Add categorize_error mapping for exit code 10 -> "config" (misc/api.func). Tweak build.func: minor pipeline formatting and change the ERR trap to capture the actual exit code and only call ensure_log_on_host/post_update on non-zero exits, preventing erroneous failure reporting. * tools: add data init and auto-reporting to tools and pve section Introduce telemetry helpers in misc/api.func: _telemetry_report_exit (reports success/failure via post_tool_to_api/post_addon_to_api) and init_tool_telemetry (reads DIAGNOSTICS, starts install timer and installs an EXIT trap to auto-report). Integrate telemetry into many tools/addon and tools/pve scripts by sourcing the remote api.func and calling init_tool_telemetry (guarded with declare -f). Also apply a minor arithmetic formatting tweak in misc/build.func for RECOVERY_ATTEMPT.
2026-02-17 16:36:20 +01:00
# Telemetry
source <(curl -fsSL https://raw.githubusercontent.com/community-scripts/ProxmoxVE/main/misc/api.func) 2>/dev/null || true
core: Execution ID & Telemetry Improvements (#12041) * fix: send telemetry BEFORE log collection in signal handlers - Swap ensure_log_on_host/post_update_to_api order in on_interrupt, on_terminate, api_exit_script, and inline SIGHUP/SIGINT/SIGTERM traps - For signal exits (>128): send telemetry immediately, then best-effort log collection - Add 2>/dev/null || true to all I/O in signal handlers to prevent SIGPIPE - Fix on_exit: exit_code=0 now reports 'done' instead of 'failed 1' - Root cause: pct pull hangs on dying containers blocked telemetry updates, leaving 595+ records stuck in 'installing' daily * feat: add execution_id to all telemetry payloads - Generate EXECUTION_ID from RANDOM_UUID in variables() - Export EXECUTION_ID to container environment - Add execution_id field to all 8 API payloads in api.func - Add execution_id to post_progress_to_api in install.func and alpine-install.func - Fallback to RANDOM_UUID when EXECUTION_ID not set (backward compat) * fix: correct telemetry type values for PVE and addon scripts - PVE scripts (tools/pve/*): change type 'tool' -> 'pve' - Addon scripts (tools/addon/*): fix 4 scripts that wrongly used 'tool' -> 'addon' (netdata, add-tailscale-lxc, add-netbird-lxc, all-templates) - api.func: post_tool_to_api sends type='pve', default fallback 'pve' - Aligns with PocketBase categories: lxc, vm, pve, addon * fix: persist diagnostics opt-in inside containers for addon telemetry - install.func + alpine-install.func: create /usr/local/community-scripts/diagnostics inside the container when DIAGNOSTICS=yes (from build.func export) - Enables addon scripts running later inside containers to find the opt-in - Update init_tool_telemetry default type from 'tool' to 'pve' * refactor: clean up diagnostics/telemetry opt-in system - diagnostics_check(): deduplicate heredoc (was 2x 22 lines), improve whiptail text with clear what/what-not collected, add telemetry + privacy links - diagnostics_menu(): better UX with current status, clear enable/disable buttons, note about existing containers - variables(): change DIAGNOSTICS default from 'yes' to 'no' (safe: no telemetry before user consents via diagnostics_check) - install.func + alpine-install.func: persist BOTH yes AND no in container so opt-out is explicit (not just missing file = no) - Fix typo 'menue' -> 'menu' in config file comments * fix: no pre-selection in telemetry dialog, link to telemetry-service README - Add --defaultno so 'No, opt out' is focused by default (user must Tab to Yes) - Change privacy link from discussions/1836 to telemetry-service#privacy--compliance * fix: use radiolist for telemetry dialog (no pre-selection) - Replace --yesno with --radiolist: user must actively SPACE-select an option - Both options start as OFF (no pre-selection) - Cancel/Exit defaults to 'no' (opt-out) * simplify: inline telemetry dialog text like other whiptail dialogs * improve: telemetry dialog with more detail, link to PRIVACY.md - Add what we collect / don't collect sections back to dialog - Link to telemetry-service/docs/PRIVACY.md instead of README anchor - Update config file comment with same link
2026-02-18 10:24:06 +01:00
declare -f init_tool_telemetry &>/dev/null && init_tool_telemetry "fstrim" "pve"
tools/pve: add data analytics / formatting / linting (#12034) * core: add progress; fix exit status Introduce post_progress_to_api() in alpine-install.func and install.func to send a lightweight, fire-and-forget telemetry ping (HTTP POST) that updates an existing telemetry record to "configuring" when DIAGNOSTICS=yes and RANDOM_UUID is set. The function is non-blocking (curl -m 5, errors ignored) and is invoked during container setup and after OS updates to signal active progress. Also adjust api_exit_script() in build.func to report success (post_update_to_api "done" "0") for cases where the script exited normally but a completion status wasn't posted, instead of reporting failure. * Safer tools.func load and improved error handling Replace process-substitution sourcing of tools.func with an explicit curl -> variable -> source via /dev/stdin, adding failure messages and a check that expected functions (e.g. fetch_and_deploy_gh_release) are present (misc/alpine-install.func, misc/install.func). Add categorize_error mapping for exit code 10 -> "config" (misc/api.func). Tweak build.func: minor pipeline formatting and change the ERR trap to capture the actual exit code and only call ensure_log_on_host/post_update on non-zero exits, preventing erroneous failure reporting. * tools: add data init and auto-reporting to tools and pve section Introduce telemetry helpers in misc/api.func: _telemetry_report_exit (reports success/failure via post_tool_to_api/post_addon_to_api) and init_tool_telemetry (reads DIAGNOSTICS, starts install timer and installs an EXIT trap to auto-report). Integrate telemetry into many tools/addon and tools/pve scripts by sourcing the remote api.func and calling init_tool_telemetry (guarded with declare -f). Also apply a minor arithmetic formatting tweak in misc/build.func for RECOVERY_ATTEMPT.
2026-02-17 16:36:20 +01:00
LOGFILE="/var/log/fstrim.log"
touch "$LOGFILE"
chmod 600 "$LOGFILE"
echo -e "\n----- $(date '+%Y-%m-%d %H:%M:%S') | fstrim Run by $(whoami) on $(hostname) -----" >>"$LOGFILE"
2024-02-15 23:04:55 -05:00
header_info
echo "Loading..."
whiptail --backtitle "Proxmox VE Helper Scripts" \
--title "About fstrim (LXC)" \
--msgbox "The 'fstrim' command releases unused blocks back to the storage device. This only makes sense for containers on SSD, NVMe, Thin-LVM, or storage with discard/TRIM support.\n\nIf your root filesystem or container disks are on classic HDDs, thick LVM, or unsupported storage types, running fstrim will have no effect.\n\nRecommended:\n- Use fstrim only on SSD, NVMe, or thin-provisioned storage with discard enabled.\n- For ZFS, ensure 'autotrim=on' is set on your pool.\n" 16 88
2024-02-15 23:04:55 -05:00
ROOT_FS=$(df -Th "/" | awk 'NR==2 {print $2}')
if [ "$ROOT_FS" != "ext4" ]; then
whiptail --backtitle "Proxmox VE Helper Scripts" \
--title "Warning" \
--yesno "Root filesystem is not ext4 ($ROOT_FS).\nContinue anyway?" 12 80 || exit 1
2024-02-15 23:04:55 -05:00
fi
NODE=$(hostname)
EXCLUDE_MENU=()
STOPPED_MENU=()
MAX_NAME_LEN=0
MAX_STAT_LEN=0
# Build arrays with one pct list
mapfile -t CTLINES < <(pct list | awk 'NR>1')
for LINE in "${CTLINES[@]}"; do
CTID=$(awk '{print $1}' <<<"$LINE")
STATUS=$(awk '{print $2}' <<<"$LINE")
NAME=$(awk '{print $3}' <<<"$LINE")
((${#NAME} > MAX_NAME_LEN)) && MAX_NAME_LEN=${#NAME}
((${#STATUS} > MAX_STAT_LEN)) && MAX_STAT_LEN=${#STATUS}
done
FMT="%-${MAX_NAME_LEN}s | %-${MAX_STAT_LEN}s"
for LINE in "${CTLINES[@]}"; do
CTID=$(awk '{print $1}' <<<"$LINE")
STATUS=$(awk '{print $2}' <<<"$LINE")
NAME=$(awk '{print $3}' <<<"$LINE")
DESC=$(printf "$FMT" "$NAME" "$STATUS")
EXCLUDE_MENU+=("$CTID" "$DESC" "OFF")
if [[ "$STATUS" == "stopped" ]]; then
STOPPED_MENU+=("$CTID" "$DESC" "OFF")
fi
done
excluded_containers_raw=$(whiptail --backtitle "Proxmox VE Helper Scripts" \
--title "Containers on $NODE" \
--checklist "\nSelect containers to skip from trimming:\n" \
20 $((MAX_NAME_LEN + MAX_STAT_LEN + 20)) 12 "${EXCLUDE_MENU[@]}" 3>&1 1>&2 2>&3)
[ $? -ne 0 ] && exit
read -ra EXCLUDED <<<$(echo "$excluded_containers_raw" | tr -d '"')
TO_START=()
if [ ${#STOPPED_MENU[@]} -gt 0 ]; then
for ((i = 0; i < ${#STOPPED_MENU[@]}; i += 3)); do
CTID="${STOPPED_MENU[i]}"
DESC="${STOPPED_MENU[i + 1]}"
if [[ " ${EXCLUDED[*]} " =~ " $CTID " ]]; then
continue
fi
header_info
echo -e "${BL}[Info]${GN} Container $CTID ($DESC) is currently stopped.${CL}"
read -rp "Temporarily start for fstrim? [y/N]: " answer
if [[ "$answer" =~ ^[Yy]$ ]]; then
TO_START+=("$CTID")
fi
done
fi
declare -A WAS_STOPPED
for ct in "${TO_START[@]}"; do
WAS_STOPPED["$ct"]=1
done
2024-02-15 23:04:55 -05:00
function trim_container() {
local container="$1"
local name="$2"
2024-02-15 23:04:55 -05:00
header_info
2024-02-19 13:13:44 -05:00
echo -e "${BL}[Info]${GN} Trimming ${BL}$container${CL} \n"
local before_trim after_trim
local lv_name="vm-${container}-disk-0"
if lvs --noheadings -o lv_name 2>/dev/null | grep -qw "$lv_name"; then
before_trim=$(lvs --noheadings -o lv_name,data_percent 2>/dev/null | awk -v ctid="$lv_name" '$1 == ctid {gsub(/%/, "", $2); print $2}')
[[ -n "$before_trim" ]] && echo -e "${RD}Data before trim $before_trim%${CL}" || echo -e "${RD}Data before trim: not available${CL}"
else
before_trim=""
echo -e "${RD}Data before trim: not available (non-LVM storage)${CL}"
fi
local fstrim_output
fstrim_output=$(pct fstrim "$container" 2>&1)
if echo "$fstrim_output" | grep -qi "not supported"; then
echo -e "${RD}fstrim isnt supported on this storage!${CL}"
elif echo "$fstrim_output" | grep -Eq '([0-9]+(\.[0-9]+)?\s*[KMGT]?B)'; then
echo -e "${GN}fstrim result: $fstrim_output${CL}"
else
echo -e "${RD}fstrim result: $fstrim_output${CL}"
fi
if lvs --noheadings -o lv_name 2>/dev/null | grep -qw "$lv_name"; then
after_trim=$(lvs --noheadings -o lv_name,data_percent 2>/dev/null | awk -v ctid="$lv_name" '$1 == ctid {gsub(/%/, "", $2); print $2}')
[[ -n "$after_trim" ]] && echo -e "${GN}Data after trim $after_trim%${CL}" || echo -e "${GN}Data after trim: not available${CL}"
else
after_trim=""
echo -e "${GN}Data after trim: not available (non-LVM storage)${CL}"
fi
# Logging
echo "$(date '+%Y-%m-%d %H:%M:%S') | CTID=$container | Name=$name | Before=${before_trim:-N/A}% | After=${after_trim:-N/A}% | fstrim: $fstrim_output" >>"$LOGFILE"
sleep 0.5
2024-02-15 23:04:55 -05:00
}
for LINE in "${CTLINES[@]}"; do
CTID=$(awk '{print $1}' <<<"$LINE")
STATUS=$(awk '{print $2}' <<<"$LINE")
NAME=$(awk '{print $3}' <<<"$LINE")
if [[ " ${EXCLUDED[*]} " =~ " $CTID " ]]; then
2024-02-15 23:04:55 -05:00
header_info
echo -e "${BL}[Info]${GN} Skipping $CTID ($NAME, excluded)${CL}"
sleep 0.5
continue
fi
if pct config "$CTID" | grep -q "template:"; then
header_info
echo -e "${BL}[Info]${GN} Skipping $CTID ($NAME, template)${CL}\n"
sleep 0.5
continue
fi
if [[ "$STATUS" != "running" ]]; then
if [[ -n "${WAS_STOPPED[$CTID]:-}" ]]; then
2024-02-15 23:04:55 -05:00
header_info
echo -e "${BL}[Info]${GN} Starting $CTID ($NAME) for trim...${CL}"
pct start "$CTID"
sleep 2
else
header_info
echo -e "${BL}[Info]${GN} Skipping $CTID ($NAME, not running, not selected)${CL}"
sleep 0.5
2024-02-15 23:04:55 -05:00
continue
fi
fi
trim_container "$CTID" "$NAME"
if [[ -n "${WAS_STOPPED[$CTID]:-}" ]]; then
read -rp "Stop LXC $CTID ($NAME) again after trim? [Y/n]: " answer
if [[ ! "$answer" =~ ^[Nn]$ ]]; then
header_info
echo -e "${BL}[Info]${GN} Stopping $CTID ($NAME) again...${CL}"
pct stop "$CTID"
sleep 1
else
header_info
echo -e "${BL}[Info]${GN} Leaving $CTID ($NAME) running as requested.${CL}"
sleep 1
fi
2024-02-15 23:04:55 -05:00
fi
done
header_info
echo -e "${GN}Finished, LXC Containers Trimmed.${CL} \n"
echo -e "${BL}If you want to see the complete log: cat $LOGFILE${CL}"
exit 0