504 gateway time-out error after reboot

[lnoferin]

My YunoHost server

Hardware: Raspberry Pi at home
YunoHost version:
I have access to my server : Through SSH | through the webadmin | direct access via keyboard / screen | …
Are you in a special context or did you perform some particular tweaking on your YunoHost instance ? : no

Description of my issue

I needed to reboot after a kernel upgrade coming from debian and after the reboot some services are not working anymore, nextcloud, lufi, rainloop.

If I try to connect to nextcloud, for example I receive the 504 gateway time-out error.

I thought it was due to a problem with php7.3-fpm and I looked in the log.

In the nginx error log I see some messages like this:

2021/07/29 23:55:05 [error] 1802#1802: *430 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 95.234.135.22, server: cybervalley.org, request: "GET /nextcloud/apps/photos/service-worker.js HTTP/2.0", upstream: "fastcgi://unix:/var/run/php/php7.3-fpm-nextcloud.sock", host: "cybervalley.org"

And in the /var/log/php7.3-fpm.log other messages like:

[29-Jul-2021 23:38:24] WARNING: [pool nextcloud] server reached pm.max_children setting (16), consider raising it

I tried to raise the pm.max_children setting (asking why now) but nothing changed.

[ljf]

How many ram on your rpi ?

[lnoferin]

I have 8Gb.

root@cybervalley:~# free -h
          total        used        free      shared  buff/cache   available
Mem:          7.8Gi       1.9Gi       4.4Gi       191Mi       1.4Gi       5.6Gi
Swap:            0B          0B          0B

[lnoferin]

I think is kind a problem for the kernel because the error is propagating over other services, like pleroma, ttrss, with 504 or 502 error.

I saw a similar report here 502 error nginx web server - Raspberry Pi Forums

[lnoferin]

Writing for myself, I will be always right!

I found the problem is in the new kernel because I downgraded it and now everything is working normally.

Now I cannot make more investigations because I am far from my server but I think the real problem is in the /boot/config file, at least reading here shdci prepare_data and swiotlb buffer is full crash system during dd operation on external usb ssd storage · Issue #4490 · raspberrypi/linux · GitHub

Currently my /boot/config is:

# Enable audio (loads snd_bcm2835)
dtparam=audio=on

[pi4]
# Enable DRM VC4 V3D driver on top of the dispmanx display stack
dtoverlay=vc4-fkms-v3d
max_framebuffers=2

[all]
#dtoverlay=vc4-fkms-v3d
gpu_mem=16
dtoverlay=pi3-disable-bt
dtoverlay=pi3-disable-wifi
dtoverlay=disable-wifi
dtoverlay=disable-bt

And as I can understand I need to add cma-256 or remove the gpu_mem if I want to use the new kernel