UNRAID - Starting PHP-FPM stuck in loop when restarting

[tiredofit]

UNRAID - Starting PHP-FPM stuck in loop when restarting

[tiredofit]

It looks as if your system only starts and stops a container, as opposed to killing it completely before restart. Can you try to find a way to stop (and remove) it from the container list, so that the container can have a "hard" start? This is one of the gotchas of some of these multi process images I build.

[BenPicard]

It looks as if your system only starts and stops a container, as opposed to killing it completely before restart. Can you try to find a way to stop (and remove) it from the container list, so that the container can have a "hard" start? This is one of the gotchas of some of these multi process images I build.

Thank you for the quick reply. There are not a lot of option possible from the Unraid UI. Maybe this is something related to the way unraid and dockers works? I have been running dockers in Unraid since 3-4 years and this is the first time that I got this error.

[tiredofit]

It's been a long time since we have seen this type of error to be honest.
I don't know much about Unraid, but it seems then if you are clicking stop is doesn't fully remove the container from the "active" container list.

If you had access to a console of unraid, you would be able to see this with by typing docker container ls and seeing that your container was still available, yet stopped.

I really only have experience with docker command line and docker-compose, When I type docker-compose down it stops and destroys the container.

You may need to fiddle around with some settings or try different scenarios - this it eh log file that jumps out at me:

[cont-init.d] 00-startup: executing...
2021-11-10-10:03:36 �[100m[NOTICE]�[49m ** [container-init] Detected Container that has been restarted - Cleaning '/tmp/state' files

That shouldn't theoretically appear if it is a "fresh" or "hard" start.
I'm certain there are other unraid users using this image as its quite popular, lets see if someone can help us? I will change the topic.

[tiredofit]

One thing you can do as well would be to go inside the container as it seems this is more a PHP issue than a freescout issue and manually type php-fpm and see what the output is?

[BenPicard]

One thing you can do as well would be to go inside the container as it seems this is more a PHP issue than a freescout issue and manually type php-fpm and see what the output is?

/ # php-fpm
[10-Nov-2021 14:56:31] ALERT: [pool www] user has not been defined
[10-Nov-2021 14:56:31] ERROR: failed to post process the configuration
[10-Nov-2021 14:56:31] ERROR: FPM initialization failed

[BenPicard]

It's been a long time since we have seen this type of error to be honest. I don't know much about Unraid, but it seems then if you are clicking stop is doesn't fully remove the container from the "active" container list.

If you had access to a console of unraid, you would be able to see this with by typing docker container ls and seeing that your container was still available, yet stopped.

I really only have experience with docker command line and docker-compose, When I type docker-compose down it stops and destroys the container.

You may need to fiddle around with some settings or try different scenarios - this it eh log file that jumps out at me:

[cont-init.d] 00-startup: executing...
2021-11-10-10:03:36 �[100m[NOTICE]�[49m ** [container-init] Detected Container that has been restarted - Cleaning '/tmp/state' files

That shouldn't theoretically appear if it is a "fresh" or "hard" start. I'm certain there are other unraid users using this image as its quite popular, lets see if someone can help us? I will change the topic.

Not sure if it's a useful information.

[tiredofit]

Something is coming up here which sounds like exactly what you are experiencing: tiredofit/docker-fusiondirectory#47

Can you output the contents of your /etc/php*/php-fpm.conf file for me?

[BenPicard]

Something is coming up here which sounds like exactly what you are experiencing: tiredofit/docker-fusiondirectory#47

Can you output the contents of your /etc/php*/php-fpm.conf file for me?

`rror_log = /www/logs/php-fpm/error.log
log_level = notice
log_limit = 2048

[www]
listen = 0.0.0.0:9000
listen.owner = nginx
listen.group = www-data

access.log = /www/logs/php-fpm/access.log

pm = dynamic
pm.max_children = 75
pm.start_servers = 2
pm.min_spare_servers = 1
pm.max_spare_servers = 3
pm.status_path = /php-fpm_status
ping.path = /ping

pm.process_idle_timeout = 10s
pm.max_requests = 500
php_admin_value[max_execution_time] = 180
php_admin_value[max_input_nesting_level] = 256
php_admin_value[max_input_vars] = 10000
php_admin_value[memory_limit] = 128M
php_admin_value[openssl.cafile] = /etc/ssl/certs/ca-certificates.crt
php_admin_value[openssl.capath] = /etc/ssl/certs
php_admin_value[output_buffering] = 0
php_admin_value[post_max_size] = 2G
php_admin_value[upload_max_filesize] = 2G
php_flag[display_errors] = on
catch_workers_output = yes

env[PATH] = /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
env[TMP] = /tmp
env[TMPDIR] = /tmp
env[TEMP] = /tmp
access.format='%{%Y-%m-%dT%H:%M:%S%z}T %{HTTP_X_FORWARDED_FOR}e %R %u %s %m "%{REQUEST_URI}e" %{HTTP_X_REQUEST_ID}e %{SERVER_PROTOCOL}e %l "%{HTTP_REFERE
~
~
~
~
~`

[BenPicard]

@tiredofit did you change something in the last update? it seems to be fixed now !!

[tiredofit]

:) I did! It just so happens that the other issue that popped up in fusiondirectory brought a change to light that was made about 1.5 months ago only surfacing now. Definitely related to containers not fully stopping and being destroyed,so I bought a few safety nets in place to avoid a duplication of configuration settings which was causing the failure. Glad it is working for you now and thanks for your patience and your submissions.

[BenPicard]

:) I did! It just so happens that the other issue that popped up in fusiondirectory brought a change to light that was made about 1.5 months ago only surfacing now. Definitely related to containers not fully stopping and being destroyed,so I bought a few safety nets in place to avoid a duplication of configuration settings which was causing the failure. Glad it is working for you now and thanks for your patience and your submissions.

Thank you so much for the quick support