POST causes segfault (502 'upstream prematurely closed connection')

[Maxim Dounin]

Hello!

On Tue, Aug 04, 2009 at 07:32:12AM -0400, miradev wrote:

> I'm stuck trying to find the culprit for an apache segfault.
> 
> I have nginx set up on one server for 2 types of site; 'normal' apache sites and wordpress fcgi sites.
> The normal apache sites have started to fail on file uploads via the html file form field.

[...]

> The nginx access log file shows a 502 response, and Apache doesn't register a hit.
> 
> Nginx error log: "upstream prematurely closed connection while reading response header from upstream"]
> Apache log: "  child pid 5566 exit signal Segmentation fault (11)"
> 
> nginx -V
> nginx version: nginx/0.7.59
> built by gcc 4.1.2 20080704 (Red Hat 4.1.2-44)
> configure arguments: --with-http_realip_module
> 
> 
> Any suggestions on where to track down this error?

As Apache dies - it's perfectly normal for nginx to show 502 
error.  Try obtaining Apache coredump to find out where it fails.

But probably the only part of the investigation relevant to this 
list is what triggers problem and possible workarounds (if any).

Maxim Dounin