🐛 [webtrees] Watchdog restarts a healthy 2.2.1 installation

[arsenyspb]

Description

When I have "watchdog" enabled, with mariadb clean Out Of The Box Experience (OOBE) installation (without even user created) I see perfectly normal 2.2.1 in a reboot loop, - comes online, exposes a webpage, logs few port 80 health checks, then reboots.

Reproduction steps

Raspberry Pi 4 (rpi4-64)
Core 2024.12.5
Supervisor 2024.12.0
Operating System 14.1
Frontend 20241127.8

Addon Logs

...snip - end of the previous successful boot...
[NV_INIT] Starting Apache
webtrees:80 127.0.0.1 - - [01/Jan/2025:17:30:19 +0800] "GET / HTTP/1.0" 400 618 "-" "-"
webtrees:80 127.0.0.1 - - [01/Jan/2025:17:30:49 +0800] "GET / HTTP/1.0" 400 618 "-" "-"
webtrees:80 127.0.0.1 - - [01/Jan/2025:17:31:19 +0800] "GET / HTTP/1.0" 400 618 "-" "-"
Starting...
/etc/cont-init.d/00-banner.sh: executing
-----------------------------------------------------------
 Add-on: Webtrees
 web's leading on-line collaborative genealogy application
-----------------------------------------------------------
 Add-on version: 2.2.1
 You are running the latest version of this add-on.
 System: Home Assistant OS 14.1  (aarch64 / raspberrypi4-64)
 Home Assistant Core: 2024.12.5
 Home Assistant Supervisor: 2024.12.0
-----------------------------------------------------------
 Please, share the above information when looking for help
 or support in, e.g., GitHub, forums
-----------------------------------------------------------
 Provided by: https://github.com/alexbelgium/hassio-addons 
-----------------------------------------------------------
/etc/cont-init.d/00-global_var.sh: executing
...snip - new boot...

Architecture

aarch64

OS

HAos

[alexbelgium]

Ahah love the OOBE definition ;) thanks for reporting I'll disable temporarily the healthcheck

[alexbelgium]

BTW I've pushed a new add-on for grampsweb! A nice alternative that is even easier to setup

[alexbelgium]

Actually there is no healthcheck instruction on the dockerfile! Not sure what causes the restarts

[alexbelgium]

It was upstream. I've implemented my own system hope it will be better

[alexbelgium]

Hi, I've pushed a new version with code refactor allowing for data folder change, please let me know how it goes!

[arsenyspb]

Testing on 2.2.1-4 with watchdog toggled on, so far no restarts.

[arsenyspb]

I have "watchdog" on in Hassio and zero reboots on 2.2.1-4 for 24+ hours, I think we can close this, thanks for patching the behaviour!

Just curious: is there a heartbeat implemented at all, one way or another?..

[alexbelgium]

Great news ! Actually I put in some effort to actually hide the heartbeat to avoid spamming the log

hassio-addons/webtrees/Dockerfile

Lines 131 to 143 in 31216ba

	RUN \
	# Handle Apache configuration
	if [ -d /etc/apache2/sites-available ]; then \
	for file in /etc/apache2/sites-*/*.conf; do \
	sed -i '/<VirtualHost/a \ \n # Match requests with the custom User-Agent "HealthCheck" \n SetEnvIf User-Agent "HealthCheck" dontlog \n # Exclude matching requests from access logs \n CustomLog ${APACHE_LOG_DIR}/access.log combined env=!dontlog' "$file"; \
	done; \
	fi && \
	\
	# Handle Nginx configuration
	if [ -f /etc/nginx/nginx.conf ]; then \
	awk '/http \{/{print; print "map $http_user_agent $dontlog {\n default 0;\n \"~*HealthCheck\" 1;\n}\naccess_log /var/log/nginx/access.log combined if=$dontlog;"; next}1' /etc/nginx/nginx.conf > /etc/nginx/nginx.conf.new && \
	mv /etc/nginx/nginx.conf.new /etc/nginx/nginx.conf; \
	fi

. Perhaps I could make the hiding optional if it is useful to you