Stange problem checking https

[Karl DeBisschop]

On Mon, 2003-02-03 at 12:43, DM wrote:
> > -----Original Message-----
> > From: Karl DeBisschop [mailto:karl at debisschop.net]
> > Sent: Sunday, February 02, 2003 10:23 AM
> > To: DM
> > Cc: nagios-users at lists.sourceforge.net
> > Subject: Re: [Nagios-users] Stange problem checking https
> > 
> > 
> > On Sat, 2003-02-01 at 20:09, DM wrote:
> > > Hi Karl,
> > >  
> > > Unfortunately the app in question is on an intranet and not
> publicly
> > > available.
> > >
> > > With your direction, I'll be happy to perform additional
> diagnostics
> > > and send you the output.
> > 
> > I was hoping to save you the compile etc. But, alas, I cannot.
> > 
> > My first step would be to download the current CVS tree, build the
> > plugins, and try that. As I had noted, the http plugin has 
> > been revised
> > several times, and some possibly relevent code was fixed.
> 
> OK, I downloaded the current CVS tree and built the plugins.
> Unfortunately, I think the new check_http module ((nagios-plugins
> 1.3.0-beta2) 1.19 )has a bug in it which prevents the SSL option from
> working at all:
>  
> $ ./check_http -w 5 -c 10 www.verisign.com
> HTTP ok: HTTP/1.1 200 OK -   0.354 second response time |time=  0.354
> $ ./check_http -w 5 -c 10 --ssl www.verisign.com
> Segmentation Fault
>  
> gdb ./check_http
> (gdb) run -w 5 -c 10 --ssl www.verisign.com
> Starting program: /temp/nagios/libexec/./check_http -w 5 -c 10
> --ssl www.verisign.com
>  
> Program received signal SIGSEGV, Segmentation fault.
> asprintf (ptr=0x0, format=0x17290 "%s") at snprintf.c:836
> 836             *ptr = NULL;
> (gdb) backtrace
> #0  asprintf (ptr=0x0, format=0x17290 "%s") at snprintf.c:836
> #1  0x13d80 in connect_SSL () at check_http.c:923
> #2  0x13aa0 in check_http () at check_http.c:552
> #3  0x1218c in main (argc=168960, argv=0x0) at check_http.c:271
> (gdb) 
> 
> Looks like the connect_SSL function is sending a null pointer to
> asprintf, making it segfault.

I'm commiting fixes for the segfault - thanks for pointing it out.

> Since I've verified earlier on that check_http is indeed returning the
> correct value to Nagios,
> I don't think the issue here is specifically with the check_http
> plugin, but rather something
> with how Nagios deals with this particular host/service.

I missed the echo $? in your previous posts, and mistakenly concluded
that the plugin was returning a diferent status than its message
showed..

This does seem strange. I can't think of any mechanism that could cause
this.

Maybe stoppong nagios and clearing all saved states might eliminate
another possible problem.

--
Karl



-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com