Two problems: Connectivity Check fails and mod_fcgid Premature end of script headers

I was setting up a new server today, moving sites off the old server to a new server to split the load, at some point the old server stopped serving pages, I stopped to fix it, but found I could not get the Connectivity Check to pass, and I get the mod_fcgid Premature end of script headers error, I see that a lot in the log, but it never took down the site before.

CentOS 5.9
PHP 5.2.17

Everything up to date.

Using bleeding edge repo, site has been running fine for years, didn’t mess we settings, just deleted some web sites.

I have been researching this all night and not getting any closer to finding a fix, I do have Virtualmin pro running on both servers right now, same license, until I get all sites balanced, so I know which license I need to buy, don’t see why that would cause problems, but I am running on empty finding a solution to bring the server back up, I took the new server down, thinking it was the license, but that did not solve the issue, new site works fine.

Php runs fine from command line, HTML will server fine, but not php,

The log does register the error each time, so DNS seems to be working, so why Connectivity Check fails now is beyond me, it doesn’t seem that deleting web sites should do this.

Validate Virtual Servers passes, and so does Checking Configuration.

Changing to cgi gives same failure, and I have time out set to unlimited.

Virtualmin runs fine using the same url as one of the sites that will not pull up, which really makes me wonder why that is, but I looked at the config, and it seems to set aside some things for port 10000.

Any help would be great.

Howdy,

Yeah, as you mentioned, the Virtualmin license isn’t going to cause the issues you’re seeing.

I’m inclined to think you’re seeing some sort of issue with your web app itself that just began occurring, that’s causing those issues.

A problem with the database, the domain reached it’s quota, or perhaps a change that was made in the web app that’s causing a problem.

Often, switching to CGI mode can generate some additional/better error messages. However, you’re saying that after you switch to CGI, your Apache error logs for that domain continue to only say “Premature end of script headers error”?

Are any domains on that server working? Or are they all generating the same error? If it’s all domains, you may want to check server-wide things… such as making sure the disk isn’t full. You could also take a look in the system Apache log, in /var/log/httpd/error_log, to see if it gives any clues.

-Eric

The License was my conscience, I had way too many sites on one server, I will now fix that, I didn’t mean to imply that VM was causing the issue, but that I had way to many sites on it that got too much traffic, at first I did wonder about doing this using the same install script, which was my Conscience saying just buy another license for 50, and never put more sites then that on a VM, that is not how to save money, and now its hosed, I will move to another VM and restore, but spit them up, I have one site that gets hammered, and it brings the others down with it, this error goes back 6 months at least, but it never took down the server like this, I set it at Apache mod_php, it works, but now I have a few email accounts that are hosed, so its much deeper then just the PHP issues.

I see this error all over the web, everyone deals with it differently, but I did not see a fix, with the email issue, I don’t think its worth troubleshooting, every once in a while you just need to start fresh and restore.

Thanks