← Docs home · Commands

Monitoring

check is read-only and safe to run as often as you like. It never installs, builds or enables anything.

Exit codes

Code	Meaning	Typical action
0	OK	none
1	AFFECTED	run build, then php --php X.Y, or re-enable after an update
2	ERROR	investigate a broken probe, missing binary or extension
3	NO IMAGICK	this server has no Imagick for the checked PHP versions

The check command exits non-zero for AFFECTED on purpose, so Nagios, monit and cron see it as a failure.

Scope --php to the versions your sites use. Without it, check inspects every installed PHP version, and a version that was never built will keep the alert firing.

Quiet output

check -q prints one line per PHP version:

php8.3 status=ok imagick=3.8.1 imagemagick=7 heic=yes libheif=1.23.7 linkage=static-hidden decode=ok
php8.3 status=affected imagick=3.8.1 imagemagick=6 heic=yes libheif=1.17.6 linkage=n/a decode=fail

Fields are stable: status, imagick, imagemagick (major), heic, libheif, linkage, decode.

Cron: silent when OK

# /etc/cron.d/heic-fix
SHELL=/bin/bash
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin

*/30 * * * * root out=$(/opt/heic-ios18-fix/heic-fix.sh check -q --php 8.3 2>&1); rc=$?; if [ "$rc" -ne 0 ]; then printf '%s\n' "$out" | mail -s "HEIC check on $(hostname): rc=$rc" ops@example.com; fi

This is completely silent when check exits 0. If it exits 1 (AFFECTED), 2 (ERROR) or 3 (NO IMAGICK), the captured output is mailed.

A wrapper can separate the alert text:

#!/usr/bin/env bash
# /usr/local/bin/heic-check-alert
set -Eeuo pipefail

out=$(/opt/heic-ios18-fix/heic-fix.sh check -q --php 8.3 2>&1) || rc=$?
rc=${rc:-0}

if (( rc == 0 )); then
    exit 0
fi

case $rc in
    1) subject="HEIC AFFECTED on $(hostname)" ;;
    2) subject="HEIC check ERROR on $(hostname)" ;;
    3) subject="Imagick missing on $(hostname)" ;;
    *) subject="HEIC check failed ($rc) on $(hostname)" ;;
esac

printf '%s\n' "$out" | mail -s "$subject" ops@example.com
exit "$rc"

Test it manually:

/usr/local/bin/heic-check-alert; echo "exit=$?"

Monit

check program heic-fix with path "/opt/heic-ios18-fix/heic-fix.sh check -q --php 8.3"
    every 6 cycles
    if status != 0 then alert

Monit treats any non-zero exit as a failure. Add separate checks if you want different alert text for AFFECTED and ERROR:

check program heic-fix with path "/opt/heic-ios18-fix/heic-fix.sh check -q --php 8.3"
    every 6 cycles
    if status = 1 then alert
    if status = 2 then alert
    if status = 3 then alert

Zabbix / Prometheus textfile

Extract a single value:

/opt/heic-ios18-fix/heic-fix.sh check -q --php 8.3 \
  | sed -n 's/.*status=\([a-z-]*\).*/\1/p'

Map it to a number in your monitoring system (ok=0, affected=1, error=2, no-imagick=3), or use the process exit code directly.

After an update

Monitoring catches the common failure mode: a GridPane or package update restores stock imagick or upgrades the system libraries. When check flips to AFFECTED:

sudo /opt/heic-ios18-fix/heic-fix.sh enable --php 8.3
sudo /opt/heic-ios18-fix/heic-fix.sh verify --php 8.3

If verify fails, rebuild:

sudo /opt/heic-ios18-fix/heic-fix.sh build
sudo /opt/heic-ios18-fix/heic-fix.sh php --php 8.3

On Ubuntu 26.04, do not rebuild. Install libheif-plugin-libde265 instead; see Ubuntu 26.04.

WordPress-native monitor

If the site runs WordPress, the standalone HEIC Fix Monitor plugin performs the same read-only checks from the PHP-FPM request context, stores a fingerprint, and sends one alert per regression or recovery by email, Slack or JSON webhook. It also provides wp heic-monitor check --notify for system cron. The cron and monit examples above remain the right choice for servers without WordPress.

Download the installable ZIP from the plugin download page.