Facebook
Twitter
Marco G
New Pleskian
Hi to everybody,
it has been 2 days I try to find a solution about this problem in all forums but i can't still find a solution.
First of all I have to tell this is a server has 64GB RAM and 12 cores AMD with Plesk 12 and has been running for more than one year without problem (and without any reboot). It is a server that has more than 40.000 visitors per day and process many files. It has been very stable until yesterday that we had to reboot.
The day before the server suddenly started to give this errors too:
and the website was down.
So we restarted the server (took very long time to restart) and disabled cloudflare cdn (so static content will be served directly from server). After restart website do not worked. And I found a solution of someone who had the same problem and I disabled nginx:
and re-enabled and it worked, the website was running. But nginx was giving a lot of errors of read (24: Too many open files).
So following instructions from a website we edit /etc/nginx/nginx.conf and we added
worker_rlimit_nofile 30000; and 4 workers. At the same time we activated cloudflare cdn. The problem seemed solved but we had many random errors on error_log from log registers of a domain in Plesk:
Before reboot we also had sometimes this errors but there were like 20 per day, now are like 300 per day.
After that I reset the default nginx.conf and the website was still stable, so maybe was the cdn that was inactive that caused read (24: Too many open files) but is strange because I think static files are served directly thru apache.
Obviously I modify /etc/httpd/conf.d/fcgid.conf to see if something changes but it didn't worked, or maybe I mess up the last config of fcgid and the current is the one that causes the problems... this is the current fcgid.conf:
More info: the /var/log/httpd/suexec_log shows many lines like:
The website is running now, but I can't reproduce the error as user so I installed a code to error_doc/internal_server_error.html and error_doc/maintenance.html to track users with this erros, and I get some, but less than the errors displayed in the error_log. I don't know if mod_fcgid: error reading data from FastCGI server error happens users can see this error page or the page it is empty.
Another thing to notice is the CPU consumption of web apache has changed since reboot:
All of this is the information I have until now.
Thank you very much for your support
it has been 2 days I try to find a solution about this problem in all forums but i can't still find a solution.
First of all I have to tell this is a server has 64GB RAM and 12 cores AMD with Plesk 12 and has been running for more than one year without problem (and without any reboot). It is a server that has more than 40.000 visitors per day and process many files. It has been very stable until yesterday that we had to reboot.
The day before the server suddenly started to give this errors too:
Code:
mod_fcgid: can't apply process slot for /var/www/cgi-bin/cgi_wrapper/cgi_wrapperSo we restarted the server (took very long time to restart) and disabled cloudflare cdn (so static content will be served directly from server). After restart website do not worked. And I found a solution of someone who had the same problem and I disabled nginx:
Code:
/usr/local/psa/admin/sbin/nginxmng --disable
/usr/local/psa/admin/sbin/httpdmng --reconfigure-allSo following instructions from a website we edit /etc/nginx/nginx.conf and we added
worker_rlimit_nofile 30000; and 4 workers. At the same time we activated cloudflare cdn. The problem seemed solved but we had many random errors on error_log from log registers of a domain in Plesk:
Code:
[Tue Oct 14 00:47:59 2015] [warn] [client xxx.xxx.xxx.xxx] (104)Connection reset by peer: mod_fcgid: error reading data from FastCGI server, referer: http://www.domain.com/page
[Tue Oct 14 00:47:59 2015] [error] [client xxx.xxx.xxx.xxx] Premature end of script headers: index.php, referer: http://www.domain.com/pageBefore reboot we also had sometimes this errors but there were like 20 per day, now are like 300 per day.
After that I reset the default nginx.conf and the website was still stable, so maybe was the cdn that was inactive that caused read (24: Too many open files) but is strange because I think static files are served directly thru apache.
Obviously I modify /etc/httpd/conf.d/fcgid.conf to see if something changes but it didn't worked, or maybe I mess up the last config of fcgid and the current is the one that causes the problems... this is the current fcgid.conf:
Code:
<IfModule mod_fcgid.c>
<IfModule !mod_fastcgi.c>
AddHandler fcgid-script fcg fcgi fpl
</IfModule>
FcgidIPCDir /var/run/mod_fcgid/sock
FcgidProcessTableFile /var/run/mod_fcgid/fcgid_shm
FcgidIdleTimeout 300
FcgidProcessLifeTime 3600
FcgidMaxProcesses 1000
FcgidMaxProcessesPerClass 16
FcgidMinProcessesPerClass 5
FcgidConnectTimeout 300
FcgidIOTimeout 1800
FcgidInitialEnv RAILS_ENV production
FcgidIdleScanInterval 10
FcgidMaxRequestLen 1073741824
FcgidBusyTimeout 1800
FcgidIdleScanInterval 480
</IfModule>
Code:
[2015-10-15 10:48:57]: uid: (10001/user) gid: (503/503) cmd: cgi_wrapperThe website is running now, but I can't reproduce the error as user so I installed a code to error_doc/internal_server_error.html and error_doc/maintenance.html to track users with this erros, and I get some, but less than the errors displayed in the error_log. I don't know if mod_fcgid: error reading data from FastCGI server error happens users can see this error page or the page it is empty.
Another thing to notice is the CPU consumption of web apache has changed since reboot:
All of this is the information I have until now.
Thank you very much for your support
Last edited:
