httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Christopher Schultz <>
Subject Re: [users@httpd] httpd 2.2 and 2.4; 500 errors with no logs at all
Date Sun, 05 Oct 2014 14:23:35 GMT
Hash: SHA256


On 10/5/14 10:01 AM, Christopher Schultz wrote:
> All,
> Over the past week, I've had 4 separate httpd servers running 2.2 
> and 2.4 start failing with the generic "Internal Server Error" page
> and a 500 response.
> The only logs generated are the access log, which of course 
> indicates a 500-response. So, no error logs, no syslogs, no 
> nothing.
> We have a couple of redirects configured using something like 
> "RedirectMatch 301 "^/$" https://hostname/" and these are
> processed and I get a 302 response which redirects to another page
> that fails.
> All of the pages that are failing are using HTTP Basic 
> authentication with an LDAPS authentication module (mod_authz_ldap)
> to a remote server. Other httpd instances are currently
> successfully authenticating against this LDAP server, so the LDAP
> server itself doesn't appear to be a problem.
> In the other cases, an httpd restart has fixed the problem. 
> Triggering a config reload (/etc/init.d/httpd reload on my 
> RHEL-compatible VM) will allow the configuration to reload and does
> not change the situation with the errors (i.e. the errors still
> occur).
> I was able to reconfigure the VirtualHost to /not/ require LDAP 
> authentication and then I was able to access my pages as expected.
>  So, I'm fairly sure that the problem is with the LDAP 
> authentication module.
> I now have another case that is failing and I have the opportunity
>  to inspect it in its failing state. Can you anyone suggest a way
> to instrument the still-running, still-failing httpd instance to 
> figure out what the problem is?
> This particular server has the following configuration as reported
>  by apachtctl -V:
> Server version: Apache/2.2.29 (Unix) Server built:   Sep 15 2014 
> 19:41:45 Server's Module Magic Number: 20051115:36 Server loaded: 
> APR 1.5.0, APR-Util 1.4.1 Compiled using: APR 1.5.0, APR-Util
> 1.4.1 Architecture:   32-bit Server MPM:     Prefork threaded: no
> forked: yes (variable process count)
> Looking at a process list, I can see 8 processes. If I go to my 
> browser and mash the RELOAD button, the browser reports a 500 
> response each time, and when I go back and check the process list,
>  all the pids are the same, so the child processes are not crashing
>  and failing to log errors before the die or anything like that.

I might have a lead on this ... the TLS certificate used by the LDAP
server has expired. But, I have some httpd servers that are happily
using it for authentication (likely due to the server restart of each
one I've had to do).

Any idea how I might be able to test this hypothesis? I'd prefer to
nail this down and call the problem solved (by issuing a new
certificate) rather than mask the real error if there is another one.

- -chris
Version: GnuPG v1
Comment: GPGTools -


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message