• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Resolved (104)Connection reset by peer: mod_fcgid: error reading data from FastCGI server

Gareth Westwood

Basic Pleskian
Hi All,

I could really do with some help trying to get to the the bottom of some problems I am having on my VPS running 18.0.25 Update #2 on Centos 6.1.

On at least 3 of the domains hosted on my server I am getting regular "500" errors being displayed and upon checking the error logs for these sites I am seeing, for example;

[Mon Apr 06 11:55:44 2020] [warn] [client ww.xx.yy.zz] (104)Connection reset by peer: mod_fcgid: error reading data from FastCGI server, referer: https://www.domain.com/wp-admin/customize.php [Mon Apr 06 11:55:44 2020] [error] [client ww.xx.yy.zz] Premature end of script headers: admin-ajax.php, referer: https://www.domain.com/wp-admin/customize.php

Two of the sites are running wordpress 5.4 (the example above is one), the other is Joomla (not sure of the version but it's up to date), all using php 7.3.16 in "Fast CGI" mode and installed using the plesk installer. DB is MySQL 5.5.39.

The Joomla site has been running for some time and the owner has complained about the 500 errors for as long as it's been up but I assumed the issue was related to Joomla so didn't do much about it. The two worpdress sites are brand new, very little content, small DB, no addons / extensions.

From the googling that I have done it seems that this error is a fairly generic "the script didn't finish for some reason" error so I'm looking for some assistance with getting to the bottom of what is causing the issue.

Can anyone offer any pointers?
 
Please check /var/log/httpd/error_log. Do you have there something like:

Fatal Error Unable to allocate shared memory segment of 67108864 bytes: mmap: Cannot allocate memory (12)

?
 
Hi Igor,

I get the error but with a memory segment of 33554432 (32mb rather than 64mb?) so I guess it's the same problem. I think I have had issues with shared memory and this server in the past. I forget exactly what the problem was but it's related to the type of virtualisation used I think.
 
Apparently, to solve the problem, you need to add the resources of this VPS.
 
Ok where does shared memory come from?

so my VPS is in a Virtuozzo container, the problem I have had previously seems to have been the same thing but it was causing opcache errors. a cat /proc/user_beancounters shows shmpages and numiptent with fail counters. I have no access to the host and my vps provider has previously said that they are unable (unwilling?) to change the setting.

It seems that the only option available to me is to reduce the amount of memory that php is using with the "opcache.memory_consumption=" directive in the php.ini

So, two questions.
1/ How do I calculate the max memory that php can use?
2/ Is that max memory shared across all of the versions of PHP.ini or do I set them all to the same thing?
 
3/ What is the minimum requirement (or how do I find the minimum) to run a modern CMS like wordpress.

a cat /proc/user_beancounters shows shmpages and numiptent with fail counters. I have no access to the host and my vps provider has previously said that they are unable (unwilling?) to change the setting.
I would insist that your provider add this parameter on Virtuozzo node side. This is easily done with a single command in the command line.
 
Good news!

My hosting provider has come back to me and they have totally removed the limit! So far since making the change I have not seen the error logged so I have my fingers crossed.

Thanks for your help Igor.
 
Back
Top