I'm having a problem migrating a website running Owncloud from a panel running on Debian 10 to a new one running Debian 12. The settings are exactly the same and I adjusted the necessary files in Owncloud, but after accessing and increasing the number of simultaneous accesses I find it difficult to keep the website online. It starts displaying status 503 and the processor usage goes to the limit. In the site log the lines: [Mon Jun 16 10:19:59.732968 2025] [fcgid:warn] [pid 32852:tid 32955] [client 1xx.2xx.1xx.8:48661] mod_fcgid: read data timeout in 600 seconds [Mon Jun 16 10:19:59.732995 2025] [core:error] [pid 32852:tid 32955] [client 1xx.2xx.1xx.8:48661] End of script output before headers: remote.php [Mon Jun 16 10:20:50.197051 2025] [fcgid:warn] [pid 32852:tid 32924] [client 19x.1xx.2xx.51:7765] mod_fcgid: can't apply process slot for /var/www/php-fcgi-scripts/web16/.php-fcgi-starter Any ideas?
I noticed that in the generated vhost the FcgidMaxProcesses option is commented out https://www.plesk.com/kb/support/we...-for-var-www-cgi-bin-cgi_wrapper-cgi_wrapper/
If you are using Debian 12, you should use PHP-FPM mode. FastCGI is the predecessor of PHP-FPM and is no longer used.
I changed to PHP-FPM but the problem persists, right after publishing the processor is always at maximum. Is there anything else to review in the settings? Log after change: [Mon Jun 16 10:54:02.884204 2025] [ssl:error] [pid 40518:tid 40518] AH02604: Unable to configure certificate cloud.lojasamigao.com.br:443:0 for stapling [Mon Jun 16 10:54:03.039130 2025] [ssl:error] [pid 40556:tid 40556] AH02218: ssl_stapling_init_cert: no OCSP URI in certificate and no SSLStaplingForceURL set [subject: CN=cloud.lojasamigao.com.br / issuer: CN=E5,O=Let's Encrypt,C=US / serial: 06AC40D10175F9D77AA70F092A72E4404F41 / notbefore: Jun 16 11:50:34 2025 GMT / notafter: Sep 14 11:50:33 2025 GMT] [Mon Jun 16 10:54:03.039142 2025] [ssl:error] [pid 40556:tid 40556] AH02604: Unable to configure certificate cloud.lojasamigao.com.br:443:0 for stapling [Mon Jun 16 10:59:24.872971 2025] [proxy_fcgi:error] [pid 40593:tid 40654] (70007)The timeout specified has expired: [client 1xx.2xx.1xx.8:48829] AH01075: Error dispatching request to : (polling) [Mon Jun 16 10:59:34.036975 2025] [proxy_fcgi:error] [pid 40593:tid 40653] (70007)The timeout specified has expired: [client 1x.1x.2x.51:7743] AH01075: Error dispatching request to : (polling)
Apaches sends a request to the FPM Backend but it reaches a timeout. I would check the phpX.X-fpm.services on the host, especially the one that you selected for this webspace. Something is not working properly.
When I try to upload the site using the PHP I need, 7.4 in this case. It stays online for a while and then crashes. Below is the log [16-Jun-2025 09:31:03] NOTICE: Reloading in progress ... [16-Jun-2025 09:31:03] NOTICE: reloading: execvp("/usr/sbin/php-fpm7.4", {"/usr/sbin/php-fpm7.4", "--nodaemonize", "--fpm-config", "/etc/php/7.4/fpm/php-fpm.conf"}) [16-Jun-2025 09:31:03] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 09:31:03] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 09:31:03] NOTICE: using inherited socket fd=8, "/var/lib/php8.2-fpm/web11.sock" [16-Jun-2025 09:31:03] NOTICE: using inherited socket fd=8, "/var/lib/php8.2-fpm/web11.sock" [16-Jun-2025 09:31:03] NOTICE: fpm is running, pid 752 [16-Jun-2025 09:31:03] NOTICE: ready to handle connections [16-Jun-2025 09:31:03] NOTICE: systemd monitor interval set to 10000ms [16-Jun-2025 10:54:02] NOTICE: Reloading in progress ... [16-Jun-2025 10:54:02] NOTICE: reloading: execvp("/usr/sbin/php-fpm7.4", {"/usr/sbin/php-fpm7.4", "--nodaemonize", "--fpm-config", "/etc/php/7.4/fpm/php-fpm.conf"}) [16-Jun-2025 10:54:02] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 10:54:02] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 10:54:02] NOTICE: fpm is running, pid 752 [16-Jun-2025 10:54:02] NOTICE: ready to handle connections [16-Jun-2025 10:54:02] NOTICE: systemd monitor interval set to 10000ms [16-Jun-2025 10:55:00] WARNING: [pool web16] server reached max_children setting (10), consider raising it [16-Jun-2025 10:55:25] WARNING: [pool web16] server reached max_children setting (10), consider raising it [16-Jun-2025 11:15:02] NOTICE: Reloading in progress ... [16-Jun-2025 11:15:02] NOTICE: reloading: execvp("/usr/sbin/php-fpm7.4", {"/usr/sbin/php-fpm7.4", "--nodaemonize", "--fpm-config", "/etc/php/7.4/fpm/php-fpm.conf"}) [16-Jun-2025 11:15:02] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 11:15:02] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 11:15:02] NOTICE: using inherited socket fd=8, "/var/lib/php8.2-fpm/web16.sock" [16-Jun-2025 11:15:02] NOTICE: using inherited socket fd=8, "/var/lib/php8.2-fpm/web16.sock" [16-Jun-2025 11:15:02] NOTICE: fpm is running, pid 752 [16-Jun-2025 11:15:02] NOTICE: ready to handle connections [16-Jun-2025 11:15:02] NOTICE: systemd monitor interval set to 10000ms [16-Jun-2025 11:21:02] NOTICE: Reloading in progress ... [16-Jun-2025 11:21:02] NOTICE: reloading: execvp("/usr/sbin/php-fpm7.4", {"/usr/sbin/php-fpm7.4", "--nodaemonize", "--fpm-config", "/etc/php/7.4/fpm/php-fpm.conf"}) [16-Jun-2025 11:21:03] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 11:21:03] NOTICE: using inherited socket fd=7, "/run/php/php7.4-fpm.sock" [16-Jun-2025 11:21:03] NOTICE: fpm is running, pid 752 [16-Jun-2025 11:21:03] NOTICE: ready to handle connections [16-Jun-2025 11:21:03] NOTICE: systemd monitor interval set to 10000ms [16-Jun-2025 11:22:31] WARNING: [pool web16] server reached max_children setting (10), consider raising it
/etc/php/7.4/fpm/pool.d/web16.conf [web16] listen = /var/lib/php8.2-fpm/web16.sock listen.owner = web16 listen.group = www-data listen.mode = 0660 user = web16 group = client6 pm = ondemand pm.max_children = 10 pm.process_idle_timeout = 10s; pm.max_requests = 0 chdir = / env[HOSTNAME] = $HOSTNAME env[TMP] = /var/www/clients/client6/web16/tmp env[TMPDIR] = /var/www/clients/client6/web16/tmp env[TEMP] = /var/www/clients/client6/web16/tmp env[PATH] = /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin php_admin_value[open_basedir] = /var/www/clients/client6/web16/web:/var/www/clients/client6/web16/private:/var/www/clients/client6/web16/tmp:/var/www/cloud.lojasamigao.com.br/web:/srv/www/cloud.lojasamigao.com.br> php_admin_value[session.save_path] = /var/www/clients/client6/web16/tmp php_admin_value[upload_tmp_dir] = /var/www/clients/client6/web16/tmp php_admin_value[sendmail_path] = "/usr/sbin/sendmail -t -i -f [email protected]"
Check the error.log of the website. I guess owncloud might also need a lot of memory, so you might increase that in the global php-fpm php.ini file of that PHP version or use custom php.ini settings field of the website to increase it.
memory_limit = 128M 256, 512, or any value, set in the options tab has no effect on the old isp panel it ran with everything in default
[Mon Jun 16 13:35:37.825573 2025] [proxy_fcgi:error] [pid 61103:tid 61156] [client 190.196.246.51:7598] AH01071: Got error 'PHP message: PHP Fatal error: Maximum execution time of 120 seconds exceeded in /var/www/clients/client6/web16/web/lib/private/Log/Owncloud.php on line 146' [Mon Jun 16 13:35:38.569516 2025] [proxy_fcgi:error] [pid 61103:tid 61158] [client 190.196.246.51:7600] AH01071: Got error 'PHP message: PHP Fatal error: Maximum execution time of 120 seconds exceeded in /var/www/clients/client6/web16/web/lib/private/Log/ErrorHandler.php on line 74' [Mon Jun 16 13:35:39.825536 2025] [proxy_fcgi:error] [pid 61103:tid 61160] [client 177.221.124.8:49345] AH01071: Got error 'PHP message: PHP Fatal error: Maximum execution time of 120 seconds exceeded in /var/www/clients/client6/web16/web/lib/private/Log/Owncloud.php on line 130'
Try setting a longer execution time. Maybe ownCloud also has some debug options, so you can find out why ownCloud needs so much processing time. Most likely, you still have old paths somewhere in ownCloud or it has a cache that you must clear or something similar.
I adjusted this as well and it didn't change anything, the high RAM usage problem continues. Remember that in the old panel nothing was changed to make it work, just the .htaccess that comes with owncloud was enough, and even then I never changed anything in it.
You do not have an issue with ISPConfig or the server configuration; you have a problem with ownCloud, so you must debug ownCloud to find out why it behaves like this. As mentioned, you likely copied over some settings or cache files that contain paths for the old server, and this must be cleared in ownCloud to fix that.
There is no file cache, everything is done in the database. I have already reviewed all possible options, reviewed configuration files and nothing. The problem persists. I will install a VM and install LAMP directly, without a panel, and see how it behaves. If it works, I will keep the ISP only as a proxy for this new VM.
As suspected, the problem with the ISP was confirmed. Running on a host with Apache, php 7.4 in its default settings, without any changes or modifications. Due to the volume of data and the short deadline to deliver the functionality tomorrow, I opted for the additional host that I was 90% sure would solve my problem. I kept a copy on the ISPConfig host and I will publish it with another address and try to understand the problem better.
Maybe it will be useful for someone, configuration used in ISPconfig for proxy on the other host with Owncloud. Code: Header always set Strict-Transport-Security "max-age=15552000; includeSubDomains" ProxyRequests Off ProxyPreserveHost On ProxyReceiveBufferSize 2048 ProxyPass /.well-known/acme-challenge ! <Directory "{DOCROOT}/.well-known/acme-challenge"> Options None AllowOverride None Require all granted AddDefaultCharset off </Directory> <Proxy *> Order deny,allow AllowOverride All </Proxy> ProxyPass / http://192.168.100.11/ ProxyPassReverse / http://192.168.100.11/
@till I noticed a problem in the latest version 3.3.0p1 that did not occur in previous versions. When you set the site to PHP-FPM, you can no longer switch back to Fast-GCI. The option disappears from the menu. Is this expected behavior or is it a bug?
This is expected as FastCGI is an outdated protocol and, therefore, disabled by default in recent versions. You can still enable it in systems settings, but it makes little sense as it has no benefits over PHP-FPM and will be removed entirely from ISPConfig in the future anyway.
