I am running Plesk 11.whateveristhelatest on CentOS 5.8 on a MediaTemple DV4.
I have one site on the server presently that today had a crush of traffic. The site is using Wordpress and therefore I have PHP configured to use FastCGI so that the user can upload files and run updates through the Wordpress Dashboard. Today, however, when the site's traffic spiked after an email blast went out, all the pages on the site started throwing "No input file specified" errors. The Plesk CGI process was also completely through the roof in resource usage and the Plesk health monitor was reporting an unbelievable 64,000%+ CPU usage.
I Googled the issue and came up with this explanation and work-around for the error, which suggests it is because of a misconfiguration of FastCGI settings in the site's vhost.conf file. What perplexes me however is that, to my understanding, vhost.conf and/or httpd.conf files are all handled by Plesk, which autogenerates its own .conf files and overwrites any changes manually made by the user.
In that case, is this not a larger misconfiguration issue within Plesk and not something I myself can address? If I am mistaken, what steps can I take to resolve this issue and ensure that it does not happen again?
My temporary fix was routing PHP through Apache and not through FastCGI. That got the error to go away and brought my processor usage back down to normal levels. But now my client can no longer upload files via Wordpress using appropriate and secure file ownership and permissions settings, so it's not a permanent solution.
Any ideas would be greatly appreciated.
I have one site on the server presently that today had a crush of traffic. The site is using Wordpress and therefore I have PHP configured to use FastCGI so that the user can upload files and run updates through the Wordpress Dashboard. Today, however, when the site's traffic spiked after an email blast went out, all the pages on the site started throwing "No input file specified" errors. The Plesk CGI process was also completely through the roof in resource usage and the Plesk health monitor was reporting an unbelievable 64,000%+ CPU usage.
I Googled the issue and came up with this explanation and work-around for the error, which suggests it is because of a misconfiguration of FastCGI settings in the site's vhost.conf file. What perplexes me however is that, to my understanding, vhost.conf and/or httpd.conf files are all handled by Plesk, which autogenerates its own .conf files and overwrites any changes manually made by the user.
In that case, is this not a larger misconfiguration issue within Plesk and not something I myself can address? If I am mistaken, what steps can I take to resolve this issue and ensure that it does not happen again?
My temporary fix was routing PHP through Apache and not through FastCGI. That got the error to go away and brought my processor usage back down to normal levels. But now my client can no longer upload files via Wordpress using appropriate and secure file ownership and permissions settings, so it's not a permanent solution.
Any ideas would be greatly appreciated.
Last edited: