Cannot navigate to website

[Cantor, Scott]

On 8/14/13 3:10 PM, "Joseph2" <joe.barnett18 at hotmail.co.uk> wrote:

>I'm sorry to keep bothering you, and I don't mean to keep coming back at
>your
>suggestions, but I am almost certain it cannot be that simple. The reason
>I
>say this is because everything was working fine the day before yesterday,
>and not much has changed since then.

I am simply going by what you posted. If it's accurate, then it is that
simple.

>I will explain a little more and perhaps you will see why I am so
>puzzled. I
>had set up https (or so I thought) and had configured most of the
>shibboleth2.xml file and the metadata.xml file, and although it was not
>fully functional I could still access the site and there didn't seem to be
>any real problems. As it turned out, I had left out the chain file when
>setting up https (I don't know if that could be significant),

That has/had nothing to do with Shibboleth, your TLS configuration is not
an SP concern.

>Surely, I would have gotten this
>behaviour and the error messages before if there was something wrong with
>the configuration from the start?

I'm not making any comment either way, I'm telling you what's wrong *now*.
You seem to be unwilling to actually go look at the shibd.log or worst
case the native.log that will tell you what's wrong. So I don't know what
else to tell you. You will not find an answer without checking the log or
possibly running a tool like xmllint depending on the nature of the error.

>Could it be that the problems were there all along and
>were just disguised somehow?

If shibd was running, then it's unlikely the file was in bad enough shape
that Apache wouldn't start, but it's not totally impossible. shibd -t is a
more complete check of the file, but an XML parsing error is fundamental,
global, unavoidable, and fatal, etc. If shibd won't load now, then there's
no mystery here, the config is invalid.

I'm not going to sit here and diagnose your configuration management. If
the file isn't in fact broken, then you can demonstrate that and we'll
move on. If it is, then you can agnonize over what mistake in your
assumptions is involved or just move on and fix it.

-- Scott